The Role of Taxonomy and Ontology in Semantic Layers - Heather Hedden.pdf
Sas sme - analytics for all
1. Anticipate adversity
Identify opportunity
Improve performance
Grow your business
insights
Analytics for all
Small and midsize businesses
see big benefits in today’s
data-rich world
Big ideas for your growing business
2
Measure it, manage it, communicate it
21
Four steps to analytic success
3
Stylish and good for you
24
Sunshine, sand and strategy
8
Analytics energizes utility
cooperative’s demand forecasts
26
Small business, big data
11
Look beyond your spreadsheets
15
Is risk management a part
of your corporate culture?
29
The Wine House discovers
$400,000 in ‘lost’ inventory
19
Analytics: An overlooked
supply chain opportunity
32
3. SMBinsights
P2
Big ideas for your growing business
by Carl Farrell, Executive Vice President, SAS Americas
Every few months a story surfaces like
this one: Small business owner signs
up with a deal-of-the-day website like
Groupon and practically ends up going
out of business because the demand
overwhelms resources.
You don’t read as frequently about a
small or midsize business that does
what Oberweis Dairy did: analyze the
potential effectiveness of a deal before
signing up and then analyze the data
to determine whether the deal turned
discount shoppers into loyal customers.
In this special Insights report, we
highlight growing businesses that once
thought the only analysis they could do
was on a spreadsheet and that analytics
was something for large companies.
These businesses have gone beyond
spreadsheets to thrive and grow by
putting their data to good use (see “Look
beyond your spreadsheets,” page 15).
We explode other myths in this report,
as well, namely that “big data” and risk
management are issues restricted to big
companies (see “Small business, big
data,” page 11, and “Is risk management
part of your corporate culture?” page
29). ACCION Texas-Louisiana has been
able to speed its loan review process
while also reducing the delinquency
rate by 76 percent. And as for big data:
Whether you’re a health care consultancy or a restaurant, you can expect
the amount of data you have to deal
with to only increase, along with the
urgency to have effective, forward-thinking ways to manage it.
Strategy expert Vijay Govindarajan,
a popular speaker at SAS leadership
events, talks about business transformation in three stages, or “boxes”: managing the present, selectively forgetting the
past, and creating the future. Together,
the latter two constitute the building
blocks for the competitive future, as the
focus “must be about what a company
needs to do to sustain leadership for the
next 10 years.”
What kind of future will you build for
your business?
As head of an expanding team of more than 1,200
professionals in over 30 countries, Carl Farrell brings
his hands-on direct leadership approach, valuescentered management and dedication to building a
winning culture of excellence to SAS and its customers.
Farrell oversees all sales functions across seven vertical
US business units, in addition to leading all business
functions in Canada, Latin America and the Caribbean.
4. P3 | sas.com/smb-insights
Four steps to analytic success
Make the case and get started with analytics
Moving from gut-feel-based decision
making to an analytics-driven approach
isn’t easy – especially for small and
midsized businesses that often face IT resource constraints, limited analytical talent, tight budgets and technology gaps.
But implementing analytical solutions that
promote fact-based decision making can
be done in a lean organization – and
the investment pays off in the end.
data goes and is mysteriously transformed into “answers.” Bruce Bedford,
VP of Marketing at Oberweis Dairy,
recommends gently introducing analytics to decision makers who don’t have
a statistical background. “Take them
through a few simple examples like a
t-test or chi-square,” he says. Then show
them how it translates into a business
decision.
The following steps can help you get
started on establishing a strong analytical foundation, but don’t view them as
a simple checklist. At every step of the
way you’ll need to reiterate the message, point out examples of analytics
accomplishments in other organizations,
and generate and publicize your early
wins to keep the attention of those who
make decisions in your organization.
The benefits of a trial program can’t
be vague. Don’t propose a project that
simply cleans data for the sake of cleaning data. Instead, explain how clean
data will improve the information you
obtain from your customer data to help
with a specific goal, like reducing churn.
If you associate your data quality initiatives with your business opportunities,
you will find support.
No. 1: Seek executive buy-in
Successful adoption of analytics starts
with securing executive support. That
means getting senior managers’ buyin from the get-go. If this is a struggle,
run a pilot program to demonstrate the
benefits. The goal is to win over a key
decision maker who can serve as the
analytics champion.
A good pilot program has the potential
not to just show where the company can
earn more or reduce costs, but actually
does it, even if in a limited scope. The
Wine House chose to look at inventory
for its first analytics project – leading it
to discover $400,000 in lost inventory.
Use a “quick win” of this sort to gain the
support of an executive champion, who
then can pave the path for adoption of
analytical solutions throughout your organization. Another quick win idea: Use
analytics to create targeted marketing
campaigns that improve response rates.
One way to earn executive buy-in is
to invite them to look under the hood.
Those unfamiliar with analytics often
view solutions as a “black box” where
Culture. This is one
area where SMBs
have the advantage,
as it is certainly easier
to make a cultural
change in a smaller
organization than in
a large enterprise.
5. SMBinsights
Golfsmith, a golf retailer, increased
direct mail response rates as much as
10 to 60 percent by using analytics to
segment customers better.
If you are having difficulty settling on the
right type of project, look for examples
outside your company, maybe even outside your industry. Create an awareness
of what other small and midsize businesses are accomplishing with analytics.
If you get those with the authority and
budget to think about the potential of
analytics, the funding will materialize.
No. 2: Establish an analytics culture
People, processes, technology and
culture – which of these four are most
important for succeeding with analytics? By far, it is culture. This is one area
where SMBs have the advantage, as
it is certainly easier to make a cultural
change in a smaller organization than
in a large enterprise.
SAS Chief Researcher Pamela Prentice
wrote about how one start-up changed
its culture: “I particularly like the notion
Jayson Tripp and Brian O’Connor from
Redbox had: ‘When someone is in the
desert starving,’ Brian said, ‘if you feed
them a saltine, it tastes like steak.’ Brian,
responsible for the business intelligence
function, initially fed Jayson, in charge
of strategic planning, ‘saltines’ of
information as Jayson worked to gain
acceptance of analytics at Redbox.
Jayson took this information and boiled it
down to three PowerPoint slides showing
the financial impact to the company –
and made inroads in analytics adoption.
Getting quick wins in using analytics
is important to winning over decision
makers.”
Just three slides? Yes, because part of
establishing an analytics culture is to
be able to tell the story in a way that
everyone – from the sales clerk to the
CEO – can understand. You can’t do
that with a 70-slide presentation that
details the minutia of algorithms and
data cleansing.
To extend the analytics culture across
the company, Kelley Blue Book created an analytic center of excellence.
The resulting improvements in decision
making help the company consistently
outperform competitors. Both costs and
benefits of analytics projects are highly
visible, making the power of analytics
clear to all.
P4
6. P5 | sas.com/smb-insights
Tie analytical
outcomes to the
strategic issues of
the business. Letting
skeptics see the
effectiveness of
using analytics
whets their appetites
and gets other
people and departments interested.
“We improved agility by placing analytic center of excellence and business
team members together,” notes Kelley
Blue Book’s Vice President of Analytical
Insights, Shawn Hushman. “As needs
arise, analytic experts contribute ideas
and answer questions.”
Tie analytical outcomes to the strategic
issues of the business. Letting skeptics
see the effectiveness of using analytics
whets their appetites and gets other people and departments interested. Over
time, this can help analytics become
pervasive throughout your business.
No. 3: Figure out at what
stage your organization is
If you can do nothing more, transition
the conversation from analytics being
about a tool or product to analytics
being a component of the business
process.
In the book Competing on Analytics:
The New Science of Winning, Thomas
Davenport and Jeanne Harris list five
stages of analytic maturity:
• Stage 1: Analytically Impaired – Lack
of analytical skill or executive interest.
• Stage 2: Localized Analytics –
Uncoordinated activities or silos.
• Stage 3: Analytical Aspirations –
Good intentions with slow progress.
• Stage 4: Analytical Companies –
Widely use analytics internally.
• Stage 5: Analytical Competitors –
Use analytics as a competitive
advantage.
Understanding where you are helps
make it easier to know where you need
to go.
To move from stage one to stage two,
walk decision makers lacking a statistical background through a few simple
examples to show how analytics translates into business decisions. Start with
tactical decisions, such as key metrics
reporting or creating basic segmentation
for more targeted marketing. Show how
tactical decisions can help drive more
strategic decisions – such as entering a
new market, introducing a new product
category based on that segmentation, or
expanding production or opening a new
plant based on demand trends.
No. 4: Identify your analytical talent
SMBs usually have very few statisticians
and analysts – if any. Most staff falls in
either the “amateurs” category – those
who use spreadsheets and run queries
– or in the “semiprofessionals” category
of those who can use some basic
statistical tools and may be able to
program in SQL. Only a few SMBs
employ “professionals” who can write
their own algorithms.
Don’t let the fact that you have only
limited analytical pros deter you. Know
that the employees that tend to lead
analytical initiatives aren’t necessarily
analytical experts, but ones who understand how to pose analytical questions.
The key is to pinpoint that talent in your
organization and create processes that
enable employees to promote analytic
best practices.
Why not just craft a job description and
hire analytic talent? Because it isn’t easy
to hire someone who knows your business and its challenges. While universities
are beginning to develop degrees and
graduate programs aimed at creating
analysts who can quickly get up to
speed on an organization’s needs, it’s
not the same as hiring from a top engineering or MBA program. Homegrown
7. SMBinsights
really does work best in this situation.
Try to identify and nurture talent that is
already at your company by looking for
individuals with both strong communications skills and technical abilities, or look
for a pair of individuals that between
them have strong business knowledge,
communication skills and technical abilities. Find online training courses and
users groups for them to attend. Local
SAS users groups, for example, are
very welcoming and encouraging for
newcomers to the field.
®
Growing SMBs might also consider
creating the position of chief analytics
officer. This person should ideally have
both a technical and business background, and be able to tell the story of
analytics in order to win over reluctant
decision makers. Short of such a position, informally identifying the technical
and business users who can pose those
analytical questions and encourage
them to work together – even if they
report up through separate chains in the
organization. Even a small company
can create a virtual center of analytic
excellence, as Kelley Blue Book did, to
promote the understanding and adaptation of analytics. At a minimum, you
want the analytic talent you do have to
get to know and work with each other –
even if they are in different departments,
or different ends of the country.
Conclusion: It’s not all about technology
Overall, you cannot ignore the human
relationships and the importance of
evangelism at every step of the way.
You need to be constantly talking about
what analytics can do and what you
have seen analytics accomplish in other
organizations. There has to be a human
element tied back to every aspect of
your work, from getting wins early on
and catching the attention of those who
make decisions in your organization.
SAS Senior VP and CMO Jim Davis on
how to become an analytic organization
What is the best way to become an analytic organization? How can
you help introduce analytics into your organization? I enjoy discussing
these questions with our customers, and these are the tips I hear most
often:
1. Look for examples outside your company AND outside your industry. For
inspiration, look not only at what you can do but also look at what other
organizations have done. Create an awareness of what other industries
are accomplishing with analytics.
2. Understand your culture. Assess the personality of your organization to
determine whether top-down or bottom-up implementation will work best.
Success can happen in either scenario – but you should know before getting started which will work best for your organization.
3. Determine your baseline. While you’re assessing the culture, you should
also assess the level of analytic understanding within the organization. A
tool like the Information Evolution Model can help you determine where
your organization falls on the adoption spectrum, so you know where to
start and what to aim for next.
4. Find a starting point. Once you’ve provided examples, you need to explain how you will apply those same methods in your organization – and
what the results will be. Cross-sell and up-sell is a natural starting point for
many businesses.
5. Give them something tangible. Especially when it comes to data quality,
tangible reasons for the project are essential. You can’t propose a project
for clean data for the sake of clean data, and expect it to get funded.
Instead, you need to explain how clean data will improve the results of
your customer data and help reduce churn. If you associate your initiatives
with your business opportunities, you will find support.
ONLINE
Four steps for your first analytics project:
sas.com/smb-firstproject
Getting started with analytics webinar:
sas.com/smb-getstarted
P6
8. P7 | sas.com/smb-insights
Tap into the right tools
Implementing the right technology is essential to building and promoting
an analytical culture in small and midsized companies. Remember, a
bad experience has longer-lasting effects than a good one. So, when
selecting a business analytics solution, look for the following capabilities:
Robust visualization
Look for strong visualization capabilities
that empower business users with even
limited technical skills to: interactively
explore large amounts of data to spot
anomalies and hidden trends; build
analytical models in a point-and-click
environment to eliminate the need for
manual coding; and share and present
these results via easy-to-understand,
dynamic graphics.
Support for advanced analytics
Seek to go beyond simple query and
reporting and OLAP drill-down capabilities. The solution should support a
comprehensive set of advanced analytical techniques, including data mining,
forecasting, scenario modeling and
optimization. This puts your organization on an equal analytical footing with
larger competitors.
Prebuilt analytical models and associated task support
Look for prebuilt analytical models to
address common business issues of varying degrees of complexity. A solution
offering model assessment tools enables
users to evaluate various models and
choose the best for the task at hand, and
to deploy and monitor the models.
Suited to a range of users
Find a solution that recognizes the talent
constraints facing SMBs and supports
basic or intermediate-level modelers
and business users. Yet don’t skimp on
features for more advanced users (such
as the option to embed a home-built
algorithm). Skill sets change rapidly, and
you want to purchase the product with
the most flexibility.
Ease of use
Seek a solution that includes data
management, analytics and reporting
capabilities via familiar interfaces such
as Microsoft Office. This ensures that
users won’t be intimidated by complex,
technical-looking interfaces that hinder
users from fully using the solution’s
capabilities.
Balanced user autonomy and IT control
Choose a solution that allows business
users to work on their own, but within
a well-defined IT environment. This
helps ensure that your already-limited IT
resources are not pressured to manage
metadata, security and data integrity
requirements at multiple locations.
Modular solutions
Choose solutions that allow your
organization to purchase the functionality
or capability you need the most right
now, while making it easy to add more
as you go – without incurring expensive
integration costs.
Training and technical support
Select solutions that include appropriate
training and technical support. Look for
resources to ensure your business users
get quickly up to speed on functionality
and can access additional support as
needed.
Low total cost of ownership
Don’t choose a solution just because it
has the lowest per-user costs. Consider
additional costs you may have to pay
for that are associated with implementation, integration, training and technical
support.
9. SMBinsights
P8
Sunshine, sand and strategy
Twiddy & Co. delivers good old-fashioned hospitality using SAS
Providing exceptional customer service
has always been – and always will
be – an essential ingredient of business
success. No one knows this better than
owners of small to medium businesses,
for whom exceptional customer service
is integral to building customer loyalty
and driving business growth. Adding to
its own unique brand of customer service
in the vacation home property management industry, one North Carolina
company is using analytics to enhance
its clients’ experience by combining
good old-fashioned hospitality with deep
business and customer insight.
For more than 33 years, Twiddy &
Company has specialized in managing
a portfolio of exceptional rental vacation
properties on the northern Outer Banks
of North Carolina. Its formula for success
over the years has remained the same:
local ownership, beautiful vacation
homes and a dedicated, experienced
staff that is 100 percent committed to
old-fashioned hospitality and a personalized approach to property management.
Commitment to customer service
To help deliver on this commitment,
Twiddy uses SAS Business Analytics to
®
“What we deliver
is hospitality – that’s
our passion and
our business – and
SAS has helped us
through the logistics
to deliver it with a
higher degree of
confidence.”
Ross Twiddy, Marketing Director
10. P9 | sas.com/smb-insights
provide exceptional home management
and maintenance on behalf of owners
while ensuring repeat visits from
rental guests.
With a portfolio of more than 900 managed properties, Twiddy has implemented SAS to support the operations
side of its business, which is helping
the family-run company better manage
budgets and expenditures related to
property maintenance.
“Over time, we captured a lot of cost
data via Excel,” explains Operations
Manager Clark Twiddy. “We decided
there had to be a way to make our data
more intelligent, and to automate the
cost analysis process so that we wouldn’t
have to rely as much on manual processes to better understand what is and isn’t
working in terms of our business strategy, and where we needed to focus our
efforts. We can now track expenditures
month-to-month and year-to-year without
doing a massive data pull into Excel,
which used to take one person a couple
of days to perform.”
Reducing errors and financial losses
For example, Twiddy is using SAS to
make better decisions about which
vendors to use for what maintenance
projects, based on cost-effectiveness,
efficiency and quality of work. And it
has reduced financial losses – due to
human processing errors of contractor
invoices – by 15 percent.
“We work with about 1,200 vendors,
and we want to assign the work to these
vendors in the most effective way possible,” Clark Twiddy says. “SAS gives
us a sense of what’s working and not
working on a dynamic basis, and that’s
something we didn’t have before. We’ve
been able to reduce our invoice processing error rate and catch costs that are
far outside the average, for any given
service, before a homeowner sees the
bill. Manually looking for these things
in the past was too resource-prohibitive.
Now we have a really good sense of
cost and quality by service provider
and apply a professional approach to
resource allocation and management.
If you’re a homeowner with Twiddy and
Company, imagine the confidence you’d
have knowing that when we coordinate
service to your home, we’re doing it in
the most cost-effective and efficient way
possible, which gives your guests the
best experience to keep them happy and
coming back to vacation at your home.”
Dynamic access to information
“Another benefit for homeowners is
transparency in terms of costs,” Twiddy
continues. “When an owner calls now
to inquire about their service costs, we
can show them their expenditures and
that they are in line with market averages. The accurate, timely and transparent information provided by SAS
delivers tangible value to our clients.
Dynamic access to information is a real
competitive advantage that we can hang
our hat on.”
Using analytics, the company is also
able to forecast and streamline the
management of routine services, such
as cleaning hot tubs and pools, before
guests arrive for their vacation.
11. SMBinsights
“One of our first projects was to automate the scheduling of pool and hot tub
service,” says Chief Technology Officer
Laura Carver. “We can now pull the
scheduling information and automatically email vendors on a weekly basis.
Before, one person would be faxing
work orders on a daily basis. This has
invented time for us that we didn’t have
before – it’s providing real savings that
we can measure on the bottom line.”
Delivering hospitality
“What we deliver is hospitality – that’s
our passion and our business – and
SAS has helped us through the logistics
to deliver it with a higher degree of
confidence,” says Marketing Director
Ross Twiddy. “We could have the best
website and the best smiles at the front
counter, but if you get to your house
and the pool and hot tub hasn’t been
cleaned, we’ve failed to provide hospitality. SAS was instrumental in making
that process better.”
According to Carver, who retained Pinnacle Consulting to help with the SAS
implementation, Twiddy & Company
plans to give homeowners online access
to the system in the future, which will
allow them to look at account information – such as revenue generated and
expenses – to support decisions about
their investment properties.
Client and employee satisfaction
Not only is SAS helping Twiddy increase
client satisfaction, it’s also having a positive impact on staff.
“Smart and motivated people want to be
measured and held accountable,” says
Ross Twiddy. “Employees have bought
into the system and like it because they
can track and see the impact they’re
having on the business. If we can work
smarter, the company and our homeowners will benefit.”
P10
How to become
best-in-class with analytics
According to recent research from
Aberdeen Group, small- to mediumsized businesses that are best-in-class
with analytics realize significant
benefits, including:
• 24 percent year-over-year increase in
new customer accounts (compared to
a 12 percent industry average).
• 16 percent year-over-year reduction in
total operating costs (compared to a
12 percent industry average).
• 18 percent year-over-year increase in
operating cash flow (compared to a 6
percent industry average).
How can you become best in class?
Start by following these three tips from
Aberdeen:
• Treat data as a strategic asset.
• Expand the reach of business
analytics.
ONLINE
Learn more about SAS Business Analytics.
sas.com/businessanalytics
• Accelerate the delivery of actionable
information.
See what SAS offers the hospitality industry.
sas.com/industry/hospitality
Check out the Business Analytics Knowledge Exchange.
sas.com/smb-baexchange
Learn more about Twiddy & Company.
twiddy.com
Learn more about Pinnacle Consulting.
psiconsultants.com
ONLINE
Read The Analytical SMB:
More Data, More Users, Less Time
from Aberdeen Group.
sas.com/smb-analyticalsmb
12. P11 | sas.com/smb-insights
Small business, big data
Surprisingly not an oxymoron, many small and midsize
businesses also deal with big data issues
by Mark Troester, Global Product Marketing Manager, SAS
Is your business too small to worry about
“big data?” Have you heard the volume,
variety, velocity descriptors about big
data and thought, “Nah, that doesn’t
apply to me”?
Think again – and consider this simple
and accurate definition for big data:
When volume, velocity and variety of
data exceed an organization’s storage
or compute capacity for accurate and
timely decision making (See figure 1).
Clearly, big data is a relative term.
Every organization has a tipping point,
and most organizations – regardless of
size – will eventually reach a point where
they will have to address the volume,
variety and velocity of their data.
goals. When you tackle big data with
big analytics, you quickly realize that
big data presents an opportunity for
every organization. Big data is not just
for multinationals. And it’s definitely not
one-size-fits-all.
More importantly, every organization
has an opportunity to use big data to
its advantage – to drive accurate and
timely decisions that can materially
affect its business and organizational
Whether your revenues are $1 million or
$100 billion, knowing how to manage
and analyze data is critical to success,
as the Economist Intelligence Unit
research illustrates well in its recent study
13. SMBinsights
Big Data: Harnessing a Game-Changing
Asset. Nearly half the survey respondents who listed big data as a major
issue facing their organization reported
revenues of $500 million or less.
Among the report’s findings:
• Over the last year, 73 percent of
survey respondents say their collection
of data has increased “somewhat” or
“significantly.”
• Companies self-identified as “strategic
data managers” – those with a welldefined data management strategy
that focuses resources on collecting
and analyzing the most valuable data
– tend to financially outperform their
competition more than others – 53
percent, compared with 36 percent.
• Thirty-two percent of self-identified
“data wasters” say they lag behind
their peers on financial performance.
Only 1 percent of strategic data users
report that.
smart meter, a utility goes from collecting
one data point a month per customer
to receiving 3,000 data points for each
customer each month, while smart meters
send usage information up to four times
an hour.
One small Midwestern utility is using
smart meter data to structure conservation
programs that analyze existing usage to
forecast future use, price usage based on
demand and share that information with
customers who might decide to forestall
doing that load of wash until they can
pay for it at the nonpeak price.
A regional trucking company provides
another example. Global position satellite
technology now allows firms to track the
trucks, the merchandise – practically anything you can attach an RFID tag to. How
a company uses that information – to reroute trucks to create efficient routes, alert
customers to deliveries, and forecast and
price services – depends on the ability to
manage and analyze data effectively.
• More than half of companies report
that they expect the increased volume
of data to improve operations. The second most popular answer (respondents
could choose two): 36 percent expect
it to inform strategic decisions.
How is your industry affected?
It’s easy to think that only certain industries generate a lot of data or deal with
new data types. For instance, retailers get
a lot of SKU data and information from
their supply chains. Financial institutions
are constantly monitoring the inflow and
outflow of money. But would you think
a small, regional utility company might
have big data concerns?
As utility companies of all sizes start to
use smart meters, they can better forecast load and reduce the need to build
additional plants. But what are the data
implications of smart meters? With a
P12
The bottom line for
organizations of all
sizes: You should not
be doing less sophisticated analysis just
because you have
more data.
Volume
Terabytes
Records
• Transactions
• Tables, files
•
•
Batch
• Near time
• Real time
• Streams
•
3 Vs of
Big Data
Velocity
Structured
Unstructured
• Semistructured
• All the above
•
•
Variety
Figure 1: haracteristics of big data (source; Big Data Analytics,
C
TDWI Best Practices Report)
14. P13 | sas.com/smb-insights
Fast-growing, regional restaurant chains
are also affected by big data. If you own
one of these restaurants, what can you
do about an onslaught of negative online
reviews? Do you have the capacity to
analyze the comments made about your
restaurants on Facebook or Yelp? As the
Economist Intelligence Unit report notes,
“Each time new kinds of data are born,
so too are opportunities to learn from
them, combine them with existing data
and create new insights.”
ONLINE
Download the report Big Data:
Harnessing a Game-Changing Asset:
sas.com/smb-bigdata
Read big data blog posts:
sas.com/smb-bigdatablogs
running your business better with big data
– it is about surviving against competition.
In another example, a health care
consultancy has made the data coming
out of medical practices the focus of its
thriving business. The company collects
billing and diagnostic code data from
10,000 doctors on a daily, weekly
and monthly basis to create a virtual
clinical integration model. The physician
practices whose data is being collected
have agreed to be measured against
90 standards of care guidelines. This
allows the independent practices to meet
Federal Trade Commission guidelines
for negotiating with health plans. The
consulting company analyzes the data to
help the groups understand how well they
are meeting the guidelines and whether
they qualify for enhanced reimbursement
based on offering a more cost-effective
standard of care. It also sends them
automated information to better take care
of patients, like creating an automated
outbound calling system for pediatric
patients who were up to date on their
vaccinations.
• It’s not size that matters. While it’s interesting for the technical discussion to
focus on size, the focus should be on
business value first. Identify your business challenge or goal. Do you need
to use blog and social media data
to analyze customer churn? Do you
need to strengthen your fraud analysis
approach by mining clickstream and
other forms of content? Do you need
to analyze many data points at the
customer transaction level? Focus on
the business value so you can align
your goals with your technical and
solution approach.
More and more companies like this one
are building their business models on the
analysis of data. So it’s not just about
IT considerations – big and small
Whether you’re in health care or the
service industry, you need to start thinking about the requirements and design
for your analytics projects. As your data
grows, so do your IT requirements and –
oftentimes – the gap between the business
need and the IT infrastructure. To overcome these challenges, consider these
points:
• Think about a different kind of big.
From a design perspective, think
about the big picture. You certainly
don’t need to take a big bang
approach in terms of implementation,
but apply standard architecture principles to ensure that you don’t
box yourself in.
• Look beyond the hype. If you do much
research on big data, you’re bound to
run across a lot of articles on Hadoop.
This new software framework for big
15. SMBinsights
data is getting a lot of attention, and
it’s a great technology, but it is not a
realistic solution for small and midsize
companies. However, just because
Hadoop isn’t for you doesn’t mean
big data is irrelevant altogether.
Consider what is best for your organizational growth before you invest
purely based on price or hype.
• Analytics is the key. In most cases,
we think about using information
management technologies like data
integration and data quality to
prepare data for analytics. Although
this is certainly an important step, the
biggest differentiator will be how you
can apply analytics to determine what
to do with your organizational data,
determine which data is relevant, and
how or whether data should be stored.
• Resources are scarce. Lack of resources,
especially the right resources to
analyze big data, is critical. In the
Economist Intelligence Unit research,
lack of the right skills to manage data
effectively is among the top two challenges cited by survey respondents
(30 percent), followed closely by “We
can’t get the data to the right people
in the organization” (23 percent) and
“We don’t have the analytic skills to
know how to use the data effectively”
(22 percent). But using data doesn’t
require hiring a team. In fact, many
successful companies start by looking
internally for the people who are
always asking questions that everyone wishes they had an answer to
and pairing them with a statistician
who can help them learn to dig into
the data. In “How to Get Started
with Analytics” on page 3, we offer
suggestions on how to deploy your
company’s internal assets to take
advantage of big data.
Solving your big data problems
with robust analytics
The bottom line for organizations of
all sizes: You should not be doing less
sophisticated analysis just because you
have more data. If the size of the data
is choking your analytics, the problem
is not that you have too much data. The
problem is that you don’t have the right
analytics environment.
This new big data world is not only
about running problems faster, but about
solving problems that were not solvable
before. As data volumes grow and new
data sources continue to multiply as well,
what new big data problems do you
have? When you put the right analytics
to work on your big data problems, you
can stop thinking of big data as only a
challenge and start seeing big data as
an opportunity.
Mark Troester oversees SAS’ market strategy efforts
for information management and for the overall
CIO and IT vision. He began his career in IT and
has worked in product management and product
marketing for a number of Silicon Valley start-ups and
established software companies.
P14
16. P15 | sas.com/smb-insights
Look beyond your spreadsheets
See the big picture instead with analytics
by Ritu Jain, Global Marketing Manager, SAS
When you talk to small and midsize
businesses about analytics, you often
hear some variation of a common
response: “Oh, we do analytics. We
just do it with spreadsheets.’’
Companies use spreadsheets for many
reasons: They are easy to use, come free
bundled with productivity software, and
most people are familiar with them, so
they require no training. But spreadsheets
also have several weaknesses that can
put a midsize company at a serious
disadvantage compared with not only
larger companies, but also other companies its size.
A research report from the Aberdeen
Group notes that, while spreadsheets can
be useful, they can also lead to serious
errors when used inappropriately.
“Spreadsheets, a widely available and
familiar tool at midsized companies, are
a poor choice for strategic activities,”
said David Hatch, Aberdeen Vice President and Principal Analyst. “Our research
shows midsize businesses that rely more
1
on spreadsheets for key analysis tend to
perform poorly across a broad range of
financial and operational metrics, as
compared to top performers who have
left spreadsheets behind in favor of
analytic applications. The danger of overreliance on spreadsheets is clear.”
One major problem with spreadsheets
is their tendency to become error-ridden.
Raymond Panko’s work at the University
of Hawaii estimates that 88 percent of
audited spreadsheets have errors.1 Some
issues with spreadsheets relate to their
original design – they were what Panko
refers to as a “scratch pad(s)” meant for
simple calculations by an individual.2
Carefully designing spreadsheets for
large-scale applications (and multiple
users) didn’t occur until after the spreadsheets were already being used – some
would say abused – in that way.
The European Spreadsheet Risks Interest Group has collected public report
of companies who have endured costly
spreadsheet errors, many of which are
R
aymond R. Panko, “What We Know About Spreadsheet Errors.” Journal of End User
Computing’s special issue on Scaling Up End User Development, Volume 10, No. 2. Spring
1998, pp. 15-21. panko.shidler.hawaii.edu/My%20Publications/Whatknow.htm
2
ibid.
17. SMBinsights
P16
“ preadsheets, a widely available and familiar tool
S
at midsized companies, are a poor choice for
strategic activities.”
David Hatch, Aberdeen Vice President and Principal Analyst
SMBs, including a European spirits maker
whose stock fell 15 percent in one day
after it discovered its revenues had gone
down, not up, and a US online retailer
whose publicly traded shares fell by more
than a quarter after a spreadsheet error
was discovered and reported.
It’s not just about errors though. Spreadsheets just don’t provide the robust
data management and analysis capabilities midsize businesses need to drive
evidence-based decision making. In the
Aberdeen report, Hatch says that the
most successful midsized businesses are
much more likely to be using powerful
and relevant technologies – technologies
that can provide automation, allow for
advanced analytics capabilities (such as
data mining, forecasting and scenario
modeling), and streamline the disbursal of
business intelligence.
Some of the primary reasons why spreadsheets should not be the answer to your
analytics needs are:
• Data Integrity and Consistency –
Have you ever gone into an important
meeting where everyone pulled out
a spreadsheet with “data” and their
interpretation of it, only to find no one
was working off quite the same batch
of data? Data can be compromised inadvertently – sometimes just the simple
process of pulling data from different
systems, different formats into one
spreadsheet can cause data modification. And if this data is saved, you may
have lost the original values forever.
Having reliable data that represents
“one version of the truth” is important
to creating an analytical data-driven
culture.
• Data Volume – Even small companies
are generating much larger data volumes than ever before. Can a spreadsheet handle millions or even billions
of cells of data? There is also the data
aggregation problem. Unless all your
data is stored in a spreadsheet format,
just the simple task of integrating all
the relevant data stored in different
formats in different data sources into
one spreadsheet can cause data errors. Even if all your data is in spreadsheets, how long does it take your staff
to gather spreadsheets from different
users and attempt to create one master
file? What about manual errors that
occur when copying and pasting data
from multiple spreadsheets?
• Manual Errors – Spreadsheets are
notorious for their susceptibility to trivial
human errors. Errors in copying and
pasting data from one spreadsheet
to another, data entry errors, omissions of negative signs (especially
in financial reporting) or demand
calculations, macro errors, accidental
deletion or modification of a cell, a
row or a column. What if your user
mistakenly sorts just a few columns of
the spreadsheet, rather than whole
sheet? Can you imagine the time and
cost of working off incorrect, meaningless data? The constant effort to keep
errors from being introduced leaves
users with little time to analyze and
plan, as SMBs told CFO magazine.
• Security – One reason there is no one
version of the truth is that spreadsheets
are so easy to disseminate. While that
makes them widely used, especially
when different departments or individuals need to provide their input, it
is not necessarily a good practice, especially if people change information,
enter erroneous data before passing it
along (without renaming the file) or if
it gets into the hands of unauthorized
individuals.
Despite the drawbacks, it isn’t easy to
get everyone from IT to business unit
leaders to agree to transition away from
18. P17 | sas.com/smb-insights
spreadsheet-based analysis. Organizations often think spreadsheet alternatives
are too costly, or will be difficult for users
to adapt to. A UMB TechWeb survey
of SMBs captured that hesitancy in the
comment of one company president: “We
need something affordable from a cost
standpoint, but also from a personnel and
resource standpoint.” As the TechWeb
survey author points out, companies and
their managers know that they need business analytics, they just struggle to
get there.
One of the best ways to get the conversation started about moving away from
spreadsheets to more robust analytic
applications is by helping leaders and
business users see the benefits of change:
ONLINE
Watch this Beyond Spreadsheets Webcast
sas.com/smb-spreadsheets
• Emphasize productivity increases.
Are staffers struggling to get critical
projects done on time? Look to case
studies that show what other SMBs
have achieved in productivity by
embracing analytics. This midsize utility
sliced in half the time it took to produce
accurate reliable forecasts.
• Focus on the ability to eliminate
“hunch”-based decision making.
A small manufacturer used analytics
to create an early-warning system for
quality issues that is employed by
business users at four separate locations. The company tried doing this
with a spreadsheet, but it took
too long.
• Discuss what doesn’t have to change.
A European publisher eliminated a
tremendous amount of data manipulation work by choosing an analytical
approach, but business users can work
with the finished product in Excel.
• Talk about the bottom line.
A midsize property management
firm reduced losses 15 percent after
its analytics program helped it see
contractor-related invoice errors that its
spreadsheet approach didn’t catch.
Once interest is established, engage IT
and business leaders in seeking out new
analytics solutions. Share some of the key
capabilities that more robust solutions in
the market offer to make the transition
from spreadsheets easier – capabilities
such as strong visualization, prebuilt
models and predictive analytics. Highlight integration options such as Microsoft
add-ins that make advanced analytical
capabilities available within a spreadsheet environment seamlessly for easier
user acceptance and adoption. For more
information, check out what to look for in
a business analytics solution.
Ritu Jain is the Global Marketing Manager for
Small and Midsize Business Solutions at SAS. She
is responsible for driving the company’s marketing
strategy in the SMB space. Prior to joining SAS, Jain
held a number of leadership positions in retail supply
chain management, driving strategic planning, global
sourcing, operations and production management.
19. SMBinsights
P18
Researchers at the Tuck School
of Business at Dartmouth College
spreadsheets in an organization
created a spreadsheet risk diag-
is directly linked to the degree
nostic tool as part of its Spread-
of risk that spreadsheets pose for
sheet Engineering Project (SERP).
Assess your
spreadsheet risk
“We’ve found that the role of
that organization.’’
1. ow important are spreadsheets in
H
your organization?
4. ow often is a spreadsheet used
H
after it is developed?
_____ Not at all important
(1)
_____ Annually
(1)
_____ Somewhat important
(2)
_____ Quarterly
(2)
_____ Important
(3)
_____ Monthly
(3)
_____ Very important
(4)
_____ Once or twice per week
(4)
_____ Daily
(5)
2. hat is the size of the spreadsheet
W
models generally created?
_____ under 100 cells
(1)
_____ 101 to 1,000 cells
(2)
_____ 1,001 to 10,000 cells
(3)
_____ 10,001 to 100,000 cells
(4)
_____ over 100,000 cells
(5)
5. hat are spreadsheets used for
W
in your organization?
(Check all that apply)
_____ Analyzing data (e.g., financial, operational) (1)
_____ Determining trends and making projections (1)
_____ None
(1)
_____ 1 other person
(2)
_____ 2 – 5 other people
(4)
_____ More than 10 other people
(5)
_____ Optimization (e.g. Solver, What’s Best)
(1)
(1)
(3)
_____ 6 – 10 other people
(1)
_____ Simulation (e.g. Crystal Ball, @Risk)
3. ow many other users are there for a
H
typical spreadsheet?
_____ Statistical analysis
Quiz results:
Categories of risk
12 or below = Low Risk.
Congratulations! Your organization is
likely using spreadsheets in the way
that they were intended, but you might
be able to do more with your data if
you used analytics.
Total Score: _____________
ONLINE
Learn more about SERP:
mba.tuck.dartmouth.edu/spreadsheet
13-16 = Medium Risk.
Like looking down the precipice? Your
organization is on the edge. If more
people use the same spreadsheet, if more
cells are added or if you depend on them
for a greater range of tasks, you’ll move
into the high-risk zone. Instead of starting
another project on a spreadsheet, look at
an analytic solution.
17 or above = High Risk.
Your organization needs to immediately
re-evaluate its dependence on spreadsheets. With the size of the spreadsheets,
the number of cells and the number of
people working on them, you are at high
risk for introducing serious mistakes that
could damage your reputation.
20. P19 | sas.com/smb-insights
The Wine House discovers
$400,000 in ‘lost’ inventory
See the big picture instead with analytics
Economic times may be tough, but
Bill Knight, owner and President of
The Wine House, is toasting a 100
percent return on his investment in
SAS Business Intelligence for
Midsize Business.
The first day the SAS application was
live, the brick-and-mortar and Internet
retailer discovered 1,000 items that
hadn’t moved in more than a year.
“That’s significant cash tied up in inventory,” says Knight. “We had a huge sale
to blow it out, generating $400,000 in
capital in one weekend, and just in time,
because in today’s economy, we’d be
choking on that inventory.”
Midsize solution, rapid implementation
With annual sales of $20 million, the
30-year-old firm is the largest wine merchant in Southern California, but with no
IT department and a point-of-sale system
that could not provide inventory aging,
Knight had no way to track the age of
his extensive inventory.
Knight attended a retail technology
conference and spoke with several
vendors, but SAS was the only one with
solutions scaled for midsize businesses.
Working with a channel partner, The
Wine House had SAS up and running
within four weeks.
Aged inventory reduced by 40 percent
“The biggest benefit of SAS has been
the ability to drill down into the specifics
of our inventory,” Knight told
Information Management magazine for
a SAS product review.
With SAS, Knight now has a real-time
view into his inventory and is able to
drill down by department, supplier,
margins, price points, age – all the way
down to pinpointing slow-moving bottles
on the shelf so that he can promote them
and move them out.
“This is a tremendous benefit for a
retailer, especially in this economy, to
know exactly what’s going on in the
business in a timely way with as much
detail as needed,” says Knight.
Using SAS, The Wine House has
reduced its aged inventory by 40
percent. “Managing inventory is a crucial
balancing act, and SAS allows us to
know exactly what’s going on so that we
can move quickly,” says Knight.
21. SMBinsights
P20
“
Now we’re using SAS to know who our individual customers
are, what regions of the country our business is coming from,
and to focus on generating more international business.”
Bill Knight, Owner and President of The Wine House
Better buying decisions
SAS also allows The Wine House to offer
customers more of what they would like to
buy. “Our reports showed us early on that
people were still buying wine. They were
just buying $20 bottles instead of $100
bottles, so we took action to shift our
inventory to lower-priced merchandise,”
says Knight.
A story from Internet Retailer further demonstrates how The Wine House is using
SAS to fine-tune buying strategy: During
the holidays, the retailer pulled back from
a planned $30,000 wine purchase because analytics revealed how much of the
supplier’s product was already in stock
and hadn’t moved in a year.
“I told him we couldn’t take any more
of his inventory until we worked through
what we had,” says Knight. “We knew
we had a problem, but other than walking the floor and recalling, we couldn’t
identify this kind of information before.”
$60,000 savings in one simple step
Now that The Wine House has a handle
on inventory, Knight is focusing on his
customers. Using SAS to clean up the
customer database and purge the mailing list, The Wine House saved $60,000
in printing and postage alone, said
Knight.
“Now we’re using SAS to know who our
individual customers are, what regions of
the country our business is coming from,
and to focus on generating more international business,” says Knight.
“The next thing we’re going to do is
build a dashboard that shows us current
customers, new customers and retention rates so that if our new customer or
retention count is going down, we’ll know
right away and can do something about
it,” says Knight.
Clean customer data = outstanding
customer service
Knight’s goal is to set The Wine House
apart by providing a “wow” experience
for customers that keeps them coming
back – for example, hosting special
wine-tasting dinners for a well-segmented
group of customers, or using buying
history to reach out to customers with a
special offer on their favorite wine before
offering it to the general public.
Why should you implement an affordable, easy-to-use business intelligence
solution? We’ll give you 10 reasons:
1. Integrate data from across your
organization.
2. Provide self-service reporting and
analysis for users of all skill levels.
3. Explore data in many ways with
a simple point-and-click interface.
4. Deliver insights with business
visualization.
5. Present data in charts, graphics
and maps with an easy-to-access,
Web-based interface.
6. Get personalized information
via customized portals.
7. Monitor performance using
dashboards.
8. Reduce decision makers’ time
looking for answers and give
them more time for making
strategic decisions.
9. Easily integrate with
Microsoft Office.
ONLINE
Learn more about SAS Business
Intelligence for Midsize Business:
Watch this Beyond Spreadsheets webcast
sas.com/smb-spreadsheets
Check out SAS for retail
:
sas.com/smb-retail
10 reasons why
you need BI
10. xpand BI capabilities at your
E
own pace and budget.
22. P21 | sas.com/smb-insights
Measure it, manage it, communicate it
Improve performance by understanding your strengths and weaknesses
By Leo Sadovy, Product Marketing Manager for Performance Management, SAS
Do you know what your strengths are? What about
your weaknesses? Here’s a typical SWOT (strengths,
weaknesses, opportunities and threats) analysis you
will find for most small and midsize businesses.
External Origin
Internal Origin
Helpful
Harmful
S W
O T
Strengths
Focus
Innovative
Agile
Opportunities
Target marketing
Service and quality
Customer loyalty
Weaknesses
Order-to-cash cycle
Risk management
Market presence
Threats
Price war
Social media
Suppliers and channels
Does most of that look familiar? Do you
see your own business in this analysis?
• Your strengths are your inventiveness,
your focus on a single organizing
idea and your ability to move quickly compared with the big players.
• Your weaknesses start with cash, extend into the area of risk where you
are probably not diversified enough
to take a big hit, and also include a
limited market presence.
• Your opportunities are the opposite
side of that coin, where being small
means that there are still billions of
potential customers out there if you
can effectively target them with
the right mix of product, service
and quality.
• On the downside, your threats are
very real: You cannot survive a price
war with the big boys; you are most
likely heavily dependent on the
Internet and social media, where
your reputation is surprisingly
vulnerable; and, not being vertically
integrated, you are dependent on
supplier and channel partners.
23. SMBinsights
P22
Performance improvement comes from
taking responsive action based on the
data, the analysis and the insight.
Let’s take another look at the SWOT
diagram. It represents the concept of
performance management: managing your entire portfolio of processes,
investments, functions, goals, risks
and objectives.
But that’s only the top-down perspective.
Measuring, managing and improving
also have to occur from the inside out.
Are you a
breakthrough company?
Looking at the 12 specific elements of the
SWOT matrix, three themes emerge:
The tools you need to manage from
this holistic perspective would include:
• Market presence, target marketing,
customer loyalty and social media
seem to create an integrated marketing management cluster – what we
call customer intelligence at SAS –
and make it the No. 1 area to apply
analytics for insight.
Only 0.1 percent of businesses
exceed $250 million in annual sales.
Are you ready to join this elite crowd?
• A strategy-driven approach, to
capitalize on your strengths, mitigate
your risks and align with your single
organizing idea.
• One version of the truth: information
management and reporting that is
timely and accurate.
• Planning and forecasting capabilities,
for efficient execution of strategy
and rapid reaction to threats and
opportunities.
• Metrics, measures, scorecards,
dashboards, reports, alerts and
communication all derived from the
trusted data mentioned above. If you
can’t measure it, you can’t manage
it. If you can’t manage it, you can’t
improve it. Finally, if you can’t
communicate it, it doesn’t exist.
• The order-to-cash cycle begs for
attention to working capital management, in conjunction with inventory,
billing and receivables.
• The dependency on third parties,
vendors and outsourcers puts a premium on supply chain management.
Taken together, improvement in these
three functional focal points within the
performance management circle depends on data-driven business decisions.
They depend on analytics for insight
and action.
In Keith McFarland’s best-selling book,
The Breakthrough Company, he reveals
strategies and skills that have enabled
certain companies to flourish into established and highly profitable organizations. While these qualities can be emulated by any organization, they seem to
be particularly relevant to SMBs, which
face many distinct challenges as they attempt to grow, including:
• Effectively managing IT spending
and responsibilities.
• Sustaining revenue during growth.
• Anticipating customers’ needs.
ONLINE
Learn more from McFarland through
our free video series of his talks:
sas.com/smb-breakthrough
24. P23 | sas.com/smb-insights
Spreadsheets serve solely to collect data
and organize and report on it, and
are often ineffective even at that limited
task as complexity increases the chance
of errors from time-consuming manual
input. Spreadsheets do not provide real
analysis or insight. Real insight comes
from data discovery, from connecting
the dots, and from segmentation and
correlation by common attributes. Taking
action on these insights, in turn, depends
on promptly communicating accurate,
trusted information to decision makers in
an understandable format and context,
along with visual displays that shout out
“Here I am!” rather than hiding on row
147, column AA.
Performance improvement comes
from consistently better, data-driven
business decisions.
Performance improvement comes
from analysis and insight.
Performance improvement comes
from taking responsive action
based on the data, the analysis
and the insight.
grow and thrive in a globally competitive
market dominated by behemoths that
are targeting you in their own SWOT
analysis.
Ask yourself how you can be more
analytical in just one of the key areas
mentioned above. Make that your goal
for 2012, and start basing decisions on
those analytics. Each year after that, bring
more analytics to bear on these priority areas in your business. Before you know it,
as your threats and weaknesses become
more manageable, you can become
guided more by your strategy, strengths
and opportunities.
ONLINE
Driven by Data: The Importance of Building
a Culture of Fact-Based Decision-Making
sas.com/smb-datadriven
This holds no matter what size your
business, but is especially essential for a
small or medium-sized business trying to
Leo Sadovy handles marketing for performance
management at SAS. Before joining SAS, he spent
seven years as Vice-President of Finance for Business
Operations for a North American division of Fujitsu.
During his 13-year tenure at Fujitsu, Leo developed
and implemented the ROI model and processes used
in all internal investment decisions. Leo has an MBA in
finance and a bachelor’s degree in marketing.
25. SMBinsights
P24
“mplementing analytics is a great way for a small company
I
to become a large company. One of the most valuable
assets of any company, large or small, is its data, but you
have to analyze it.”
Bruce Bedford, Vice President of Marketing Analytics and Consumer Insight, Oberweis
Stylish and good for you
From fashion to dairy, timely customer touch points help companies grow
Most people would be hard-pressed to
find similarities between an online luxury
fashion retailer and a milk delivery business in the suburban Midwest. While
Gilt Groupe has made a name for itself
selling fashion-forward merchandise to
budget-conscious trendsetters, Oberweis
Dairy focuses on people with a yen for
getting their dairy products the old-fashioned way – delivered right to their front
doors. However, when it comes to customer service and optimizing marketing
programs, these two companies have
more in common than you might think.
Targeting the right customers with the
right offers will effectively lead to higher
response rates, improved channel
effectiveness and reduced marketing
spending. It also means fewer deleted
emails and fewer unwanted direct mail
solicitations.
Both are focused on satisfying customers
so they will remain loyal. And both use
analytics to plan, prioritize and optimize
customer communications.
Tapping into the market for thrifty
lovers of luxury
Gilt Groupe has grown its membership
mailing list by 90 times in just four years.
As the company added more members
and new merchandise to its mix, managing all the customer data in disparate
systems and formats became a huge
challenge for marketing analysts, who still
relied on manual SQL (Structured Query
Language) queries. It took them a long
time to produce reports, and the company couldn’t easily segment its customers.
Gilt Groupe recognized that it needed
a better way to gather, analyze and
report data so it could know customers
more intimately and be able to customize
How do they do it? They use marketing
optimization, which applies mathematical techniques to maximize economic
outcomes, making the most of each
individual customer communication.
For example, marketers can increase
campaign ROI by determining the right
offers for the right customers by using
what-if analysis and taking into account
things such as customer preferences,
propensity to buy, profitability, costs and
contact policies.
Gilt Groupe and Oberweis are optimizing their marketing with many of these
techniques, and their successes provide
insight into what small to medium businesses can do without hiring a team of
analysts and programmers.
26. P25 | sas.com/smb-insights
marketing more effectively. “We were
able to get a deep understanding of
our customer base through the profiling,
segmentation [and] predictive analysis
that we conduct with SAS Analytics,”
said Tamara Gruzbarg, Senior Director
of Analytics and Research. With analytics
the company can now “dig deep into all
of the behavioral patterns and understand the preferences of different customer segments.” Among its successes:
• A 10 to 20 percent lift for customers
browsing in new merchandise categories who had not purchased in those
categories.
• A 100 percent lift (for the first three
deciles) for women who shopped
at the men’s site but had not yet
purchased.
• A 20 percent increase in new
member conversion rates (customers
who join but haven’t purchased).
Gruzbarg’s counsel to other SMBs: “It
is never too early to start with analytics,
even if you don’t have full-blown capabilities right away. Simple segmentation
based on one or two key variables,
implemented at the right time, could go
a long way in helping to move the business forward.”
Reinventing home milk delivery
for a 21st century world
In Illinois, Oberweis Dairy is growing
as a regional food manufacturer and
retailer with home delivery, dairy stores
and a wholesale business. Each business model has different customer databases and information systems, but with
analytics the company can look at customers across all channels. For instance,
Oberweis linked its “Moola” customer
loyalty card, representing in-store sales,
with the home delivery customer database. The company found that it could
easily mine dairy store receipt data,
match it against loyalty card information
and select the best candidates for home
delivery sales campaigns.
It also learned that running specials on
milk sold through grocery store chains
doesn’t cannibalize from dairy store or
home delivery sales. In fact, a sevenfold
increase in sales at grocery stores during a recent promotion helped introduce
many new customers to the Oberweis
brand. And one of the side benefits of
switching to analytics is more productivity.
Top benefits of
marketing analytics
1. ncrease response rates, customer
I
loyalty and ultimately ROI by
contacting the right customers with
highly relevant offers at the right
time through the right channel.
The company was able to automate reports that previously took its analysts 20
hours a week using Excel spreadsheets.
“I think implementing analytics is a great
way for a small company to become a
large company,” says Bruce Bedford,
Vice President of Marketing Analytics
and Consumer Insight. “One of the most
valuable assets of any company, large
or small, is its data, but you have to
analyze it.”
ONLINE
Discover how Oberweis ramped up its marketing
effectiveness with analytics.
sas.com/smb-oberweiswebcast
Learn more about Gilt Groupe’s exploration of the
behavioral patterns of its shoppers.
sas.com/smb-giltwebcast
Learn how to fuel marketing effectiveness.
sas.com/smb-gilttips
2. Reduce campaign costs by targeting
customers most likely to respond.
3. Decrease attrition by accurately
predicting customers most likely to
leave and developing the right
proactive campaigns to retain them.
4. Deliver the right message by segmenting customers more effectively
and better understanding target
populations.
ONLINE
Learn more:
sas.com/smb-mktanalytics
27. SMBinsights
P26
Analytics energizes utility
cooperative’s demand forecasts
Customers save millions thanks to more accurate forecasting
The Old Dominion Electric Cooperative
(ODEC) saved utility customers millions
in its first year of using SAS Analytics to
forecast energy demand. The savings
helped the not-for-profit lower rates
four times. With better forecasts, the
cooperative hopes to continue keeping
costs low and service levels high.
ODEC provides wholesale power to 11
not-for-profit distribution cooperatives in
Virginia, Maryland and Delaware that
serve 1 million member customers in the
rural and suburban portions of those
states. “Each cooperative has unique
characteristics, its own weather and
economic drivers that affect growth,”
explains David Hamilton, Manager of
Load Forecasting. ODEC owns power
plant assets and also seeks to purchase
power. For energy purchases, the
cooperative must contract months in
advance. Bet wrong about the weather
or energy needs, and ODEC is at the
mercy of the spot market.
“If I don’t buy enough, I have to pay
whatever the market price is at the
time I need to buy. If you have excess,
you have to sell it for whatever price
you can get,” Hamilton says. “The
electric utility field is fairly unique. But
the problems we face each day are
the same as those in energy, gas or
oil.”
SAS allows Hamilton to forecast more
efficiently. This provides ODEC with
nimbleness when it comes to buying
and selling power and planning for
the future. “When you’re investing up
to $3 billion in a power plant, you
need to be sure you’re going to use it
when you build it,” Hamilton says.
SAS Forecast Server allows Hamilton’s
department to use the most sophisticated forecasting models and techniques available, including exponential
smoothing models, ARIMAX models,
unobserved components models,
28. P27 | sas.com/smb-insights
intermittent demand models and dynamic
regression – plus user-defined models.
SAS models are used to support system
analysis, hedging models, financial
forecasts, and future resources for energy
and demand. With SAS, ODEC can:
• Quickly adjust for changing
conditions. Forecasts take half the
time to build. Unforeseen changes – a
cooler summer or colder winter – can
be quickly worked into a forecast.
• Manage effectively despite the volatility that smaller energy providers are
more prone to experience.
“Utilities with large loads can stand a
lot of variants and still have a pretty
good forecast. We’re much smaller and
our variability has a propensity to be
higher,” Hamilton says.
• Understand each cooperative’s individual market while also aggregating
data for a big-picture look. Individual
market snapshots help ODEC choose
where to buy power from. An aggregate look helps plan for power needs
five, 10 or 20 years down the road.
“We couldn’t do what we do without
SAS,” Hamilton says. “There is no other
software I know of that has that amount
of flexibility and power.” And it pays
enormous dividends to ODEC’s member
customers. “We actually lowered the
rate we charge for wholesale power.
The cooperative can pass that benefit
directly along to the members who
have been struggling.”
• Factor in multiple data sources from
retail sales to population trends along
with daily weather information that
goes into such detail as wind speed
and cloud cover. SAS allows ODEC
to understand every variable in a
model and how it contributes to a
model’s results. The cooperative can
run competing models against each
other to choose the best one. It can
also screen outlier factors – like a hurricane – to avoid skewing the model.
Streamlining the forecasting process
In the past, ODEC used SAS for some
reporting but used another forecasting tool and Excel spreadsheets to
cobble together forecasts. “It was
labor-intensive, but people understood
the spreadsheets so that’s how it was
done.” Hamilton found other forecasting products lacking in capabilities and
believed from his prior use of SAS that
SAS Forecast Server would provide a
more robust solution.
Hamilton also uses SAS to answer
analytics requests from other ODEC
staff members and to look at data that
comes in from meter readings. “SAS
came in really handy for the AMI
(meter) data because the sheer volume
would outstrip any Excel application or
basic desktop application,” says John
Robinson, Business Systems Analyst.
“Our organization is not one to add
people. If we need to do another project, then I need to wear another hat.
We couldn’t answer these questions for
the organization if we didn’t have SAS,”
Hamilton adds.
Working with SAS to implement SAS
Forecast Server made the process
smooth, says Hamilton. ODEC uses
SAS partner Zencos to administer the
solution. A SAS Gold Partner, Zencos
provides SAS clients with services
required for installing, optimizing and
managing the SAS Business Analytics
Framework. “The whole team at SAS
has been so helpful,” Hamilton says.
Ultimately, Hamilton says, SAS takes
the guesswork out of interpreting
forecasts. Other solutions can tell him
that power usage is down and trending toward staying that way, but only
29. SMBinsights
“ ur organization is not one to add people. If we need to
O
do another project, then I need to wear another hat. We
couldn’t answer these questions for the organization if we
didn’t have SAS.”
David Hamilton, Manager of Load Forecasting
SAS helps him understand if that is
related to weather or the economy or
both. “If I didn’t have SAS, I probably
wouldn’t know how to do this. It’s nice
to be able to get a feel for how much
variability is in each component that
drives our sales. I know SAS has a lot
of different customers with a lot of data
that use SAS in different ways. It’s the
same for an electric utility. We have a
lot of data. We have to have a system
that can crunch large data sets, and
you can’t do any kind of analysis on
these large sets without SAS.”
Four tips for understanding the future
It has never been easy to forecast – whether it’s the customer demand
for a new product or a service, or the profit potential of a sales promotion. But in recent years, economic uncertainty and changing customer
behaviors have made the job of forecasting even more complex.
It’s not enough to just look at past trends. Today’s customer is more
price-sensitive, brand loyalty is declining, and competitive activity is
fiercer. Keep these tips in mind when starting a forecasting project:
1. Spreadsheets can’t do the job. They lack a forecasting algorithm,
require custom coding to develop forecast models (which makes
them error-prone and user-biased); and are incapable of handling
large volumes of data.
2. Shaping demand is a critical component. To do that, you need to
ONLINE
Read regular updates and avoid forecasting
pitfalls by subscribing to the blog, The Big
Forecasting Deal
blogs.sas.com/forecasting
Let’s talk forecasting
sas.com/smb-forecasting
Watch this SAS Forecasting demo in 5 minutes:
sas.com/smb-demo
Scan the QR code* with your
mobile device to view the video
“The Hype and Hyperbole
Around Forecasting.”
Generated by BeQRious.com
*Requires reader app to be installed on your mobile device.
perform “what if” analyses to assess the impact of changes in your
strategies on your demand, so that you can optimize your decisions.
3. For more accurate forecasting, you have to take into account not only
the impact of market trends and seasonality, but also the impact of
competitive activity, sales promotions, new product introductions, pricing and other causal factors on your demand.
4. In most situations, a solution that doesn’t allow for the impact
of multiple variables is of limited use. Not being able to model
the impact of multiple variables at the same time on demand or
perform “what if” analysis can result in stockouts, overages, poor
order fulfillment rate, and customer dissatisfaction.
P28
30. P29 | sas.com/smb-insights
Is risk management a part
of your corporate culture?
The profitable results of anticipating adversity and capitalizing on opportunity
by David Rogers, SAS Global Product Marketing Manager for Risk
Can analytics help drive cultural
change? Yes, especially as it relates to
managing risk. Corporate culture is the
foundation for any business. It dictates
how employees will treat customers and
one another, and it molds the kind of image and brand reputation that management desires.
Large companies have the resources
to manage risk in a holistic fashion –
through both personnel and technology.
Small and midsized companies, on the
other hand, have the same compliance
and risk issues – but without the same
resources. These companies are often
forced to use “one size fits all” solutions
that limit their capabilities or slow their
efficiency. Or they turn to consultants
who encourage an expensive dependency.
What financial SMBs need is the ability
to drive their risk culture with solutions
that assist them in keeping the process
in-house and efficient. They need the
option to tackle one issue (risk or
compliance) with the flexibility to manage growth and the addition of additional solutions – with the personnel
they already have.
In the insurance world, “it’s expensive
for small and midsized carriers to hire
full-time actuaries and statisticians, and
it can be difficult to find employees with
detailed knowledge about the spectrum
of distribution channels and customer
segments – there’s no substitute for experience,” explains Brian Scott, Managing
Partner of Triad Analytic Solutions.
“It’s also hard to attract employees to
certain locations.”
Triad works at getting SMBs to the
stage where they can take over and do
advanced analytics – without scores of
statisticians and actuaries. For that, Triad
sets up customers with flexible tools that
can manage large data sets and allow
staff to see the information in easy-to-use
formats.
“In one engagement, the client was in
the midst of a multiyear, seven-figure
project with an IT vendor, which was
designed to deliver critical data access,”
recounts Chris Hardin, also a Managing Partner with Triad. “In a matter of
months, we were able to use [a solution]
to build and query an interim database
that yielded similar, actionable analytic
pricing information. We also trained
the employees to maintain and use the
database we created.”
Credit risk assessment at the desktop
ACCION Texas-Louisiana is a nonprofit microloan provider that isn’t large
enough to support a team of credit risk
analysts. Its loan portfolio is just $26 million. ACCION provides loans to small
business owners who don’t qualify for
traditional loans. With a desktop analytics solution, the organization was able
to speed the loan review process while
also reducing the delinquency rate by
76 percent and loan restructuring rate
by 64 percent.
With help from a consultant, the nonprofit was able to create a scorecard
that halved the number of loan applications requiring underwriter review – the
remainder are automatically approved
or denied. Also, loan officers can now
prepare an application in only 30 minutes compared to four hours, and loan
approval times have plunged from two
weeks to three days.
“The scorecard gives us the information
needed to manage risk, increase our
efficiency and provide faster turnaround
31. SMBinsights
P30
“ e’d been looking for suspicious, fraud-related activity using
W
manual reports and it was very labor intensive; we were
very interested in an alternative to enhance and streamline
the process.”
Nancy Huntoon, Security and Fraud Manager and Bank Secrecy Act (BSA)
Compliance Officer, Northwest Federal Credit Union
times for our customers,” said Janie Barrera, ACCION Texas-Louisiana President
and CEO. To minimize default rates,
ACCION Texas-Louisiana evaluates 35
separate criteria to find individuals capable of repaying small loans. The ability
to quickly score applicants on its unique
criteria keeps its underwriting staff lean
and frees up loan officers to seek good
candidates. The organization needed
a solution that could be developed and
maintained by management and staff
with no programming experience.
The scorecard, managed in-house by
ACCION Texas-Louisiana, was so successful that it helped the nonprofit win
the business of underwriting loans for
14 additional microfinance organizations nationwide. Citigroup Inc. partnered with ACCION Texas-Louisiana,
purchasing up to $30 million worth of
microloans, because of its successful
scorecard-based prequalification tool
and loan management capabilities.
Forecasting capabilities for the SMB
Many financial institutions need to look
beyond individual risk. For SMB organizations factoring in broader economic
variables, such as high unemployment
rates, it has been difficult. One issue
in looking at risk is the ability to go beyond just looking at individual risk to get
a bigger picture. In California, Wescom
Credit Union is able to forecast possible
losses and enable mitigation activities,
from declining loan applications to
improving collections.
“We still want to provide members with
credit and at the same time ensure the
ongoing safety and soundness of the
credit union,’’ says David GumpertHersh, Wescom’s Vice President of
Credit Risk. “Now we are able to
measure risk more effectively than we
would by using a single attribute, such
as FICO (Fair Isaac Corporation), in
making credit decisions.” The credit
union estimates that it can work with at
least 50 percent greater accuracy when
deciding whether a loan will “perform”
or “not perform.” “Thanks to forecasting,
mitigation and strategic planning, we’ve
saved millions of dollars and been able
to improve forecast accuracy.”
Managing compliance risk on a
small budget
Risk doesn’t just exist in a loan portfolio. It also takes the form of complying
with government regulation related to
anti-money laundering. Although this is
thought of as more of an issue for international banks, credit unions, regional
and community banks are facing similar
regulatory scrutiny.
In the Washington, DC metropolitan
area, Northwest Federal Credit Union
(NWFCU) was trying to meet its antimoney laundering obligations with a
highly manual system. The five-branch
credit union used a system that generated a number of transactional reports that
contained a large volume of detail each
day and then manually combed through
each transaction to identify possible
fraudulent activity. “We’d been looking
for suspicious, fraud-related activity using
manual reports, and it was very labor
intensive; we were very interested in an
alternative to enhance and streamline
the process,” says Nancy Huntoon,
Security and Fraud Manager and Bank
Secrecy Act (BSA) Compliance Officer
at NWFCU. “Our internal system did
not generate alerts and we had daily,
weekly and monthly raw data reports
– basically, 50 to 100 pages of transactions to look over. It was very difficult to
get through them each day.”
The credit union now uses a solution that
crunches all the data and provides the
fraud and money laundering scenarios
and risk factors up front. “It’s helped us
with our time management by identifying possible fraudulent activity or transactions, which allows us to focus more
accurately on suspicious alerts. Now,
our BSA specialist can effectively review
32. P31 | sas.com/smb-insights
A catalogue of risk
With potentially hundreds of risks
that can be identified, dealing with
them may seem daunting. Let’s break
it down into more manageable
chunks and start by categorizing
various risks. Risks could be grouped
in any of a number of ways: external
and internal; controllable and uncontrollable; or insurable and uninsurable. Four alternative types include:
1. Market and price risk. There is a
risk that an increase in product
or service offering supply or an
aggressive price reduction from
competitors will force lower
prices and consequently reduce
profits.
2. Credit risk. The threat that
customers will fail to pay for
their purchases.
3. Operational risk. The potential
for loss resulting from inadequate
or failed internal strategy,
processes, people and technology, or from external events.
4. Legal risk. The financial risk from
insufficient net positive cash flow
or from exhausted capital-equity
raising or cash-borrowing capability. The risk from litigation or
regulatory authority penalties.
Read how to develop a risk
assessment map
sas.com/smb-riskmap
“ hanks to forecasting, mitigation and
T
strategic planning, we’ve saved millions
of dollars and been able to improve
forecast accuracy.”
David Gumpert-Hersh, Vice President of Credit Risk, Wescom Credit Union
the new alerts and move on to monitoring case activities on a daily basis. In
the past, it might have taken the specialist more than a day to just work through
the large volume of transactional data.”
According to Huntoon, her team monitors a wide variety of fraud scenarios
and risk factors – such as large cash
withdrawals and deposits, wire transfers, the velocity of debit card activity
and money structuring – and says the
predefined scenarios and factors in the
solution helped the credit union adapt
to the new system very quickly. To build
upon the credit union’s financial crimes
capacity, Huntoon and her team are
also building a householding process
to analyze the relationships between
member accounts, and the money that
moves between them, to identify suspicious activity.
“The flexibility [of the solution] allows
you to synthesize your analysis of transactions and prevent potential fraud,”
concludes Huntoon.
Whether the issue is credit risk, forecasting where the loan portfolio is going,
pricing insurance policies to mitigate risk
or efficiently meeting compliance regulators, SMBs do have analytical options
that are not only cost-effective but also
give companies a competitive edge.
ONLINE
Learn more about how business analytics
can help financial institutions.
sas.com/smb-financialwp
Gain insights on critical risk related issues
through the Risk Knowledge Exchange.
sas.com/smb-riskexchange
David Rogers is Global Product Marketing Manager in
Risk at SAS. He works closely with global SAS strategists,
product and program managers, and liaises with customers, partners and industry analysts to ensure that SAS
understands the developing risk management market.
His areas of expertise include delivery of enterprise risk
management solutions and architectures, and financial
services data integration and reporting.
33. SMBinsights
Analytics: An overlooked
supply chain opportunity
From product quality to customer delivery, smarter really does mean better
by Ritu Jain, Global Marketing Manager, SAS
Every time I talk to supply chain
professionals – whether in procurement,
demand planning, manufacturing or
delivery – one common theme keeps
surfacing: the need for better analytics.
But what can business analytics do to
enhance a supply chain? Supply chain
managers face tough questions every
day. Questions such as:
• Who are my best suppliers?
• How much do we as an organization spend on various commodities
and materials?
• Can we consolidate our supply
base without increasing risks to our
supply chain?
• What would be the impact of
adding another sales promotion on
product profitability?
• Which of the various marketing
strategies and sales tactics are most
profitable?
• What will be the impact of changes
in fuel prices, weather or a competitive promotion on product demand?
• How can we minimize our
inventory carrying costs without
affecting customer service?
• How can we identify product
quality issues earlier to minimize
our warranty claims?
• How much should we reserve for
warranty costs?
Business analytics helps supply chain
professionals answer these questions
by providing them with data-driven
insights into demand, supply network
vulnerabilities, operations and customer
service requirements. Leading companies have repeatedly shown that by
accessing and analyzing supply chain
data from all pertinent sources, forecast accuracy can be improved,
inventory can be optimized, emerging
issues can be detected early enough
to be addressed at the product-design
level (rather than in the field), and
fraudulent service claims can be
identified and eliminated. With
analytics, businesses get a complete
picture of their operations that enables
them to align their supply chain goals
P32
34. P33 | sas.com/smb-insights
Measuring what
really matters
The success of your supply chain
analytics depends on many factors,
but this is of utmost importance: Don’t
get caught measuring just for the sake
of measurement. The performance of
each objective can be measured in
multiple ways: by cost, effectiveness,
time dimension and so on. If you
measure everything, the result is many
meaningless and conflicting metrics
that don’t directly relate to the end
objective.
The key to choosing appropriate
metrics lies in understanding which
metrics really matter – and knowing
your key objectives. To ensure overall
improvement in supply chain performance, it is important to balance
departmental and geographic goals
with strategic, enterprisewide goals.
However, it is easier said than done.
When it comes to supply chains, most
managers aspire to achieve too many
objectives simultaneously without
accounting for inherent trade-offs. For
example, the goal of reducing supply
costs may have a negative impact on
desired product quality, lead time or
the proximity of the supply base.
We suggest getting everyone together
to align these cross-departmental
goals and ensure that operational KPIs
map to the company’s strategic goals.
Make sure you’re not measuring and
rewarding against conflicting metrics.
This can cause inefficiencies and
counterproductive decisions, including
those that lead to recalls.
with the organization’s strategic
business goals of improved profitability
and increased customer satisfaction.
Take the example of BGF Industries. A
manufacturer of high-end specialty
woven and nonwoven materials made
from glass, carbon and other strong,
heat-resistant fibers, BGF was seeing its
business change in front of its eyes. At
one time, 70 percent of the company’s
materials were used in electronics
manufacturing, such as printer circuit
boards. Today, many other industries
use the materials in products as diverse
as hot-air filtration systems, aircraft and
automotive parts, and bullet-proof vests.
As the company transitioned to more
quality-sensitive industries (aerospace,
defense), pressure intensified on product quality initiatives, which resulted in
unforeseen issues.
Though the company collected millions
of data points, it just wasn’t equipped
to analyze that data in a timely manner
to resolve production and quality issues.
Just moving the data out of its business
systems into Excel took more than 30
minutes. Manually reviewing the data
to identify root causes was just not
feasible or time-effective. Moving to a
new system that included robust
analytical capabilities helped resolve
production and quality issues quickly.
BGF was able to track and highlight
issues as material batches were run,
thus fixing the problem during the
production – not afterward. As a result,
BGF got a higher yield, less scrap
and – best of all – confidence in its
data and decision making.
So when the bottom- and top-line
benefits are so significant, why aren’t
more companies taking advantage
of analytics?
I would say it’s a combination of many
factors, ranging from limited analytical
talent, and siloed and incomplete data,
to the limitations of current technological infrastructures. But, that is not the
complete story. Industry studies show
that while companies recognize the
need for analytics, only a few are
harnessing the benefits of the available
technology even to a moderate extent.
In fact, a recent survey of more than 200
supply chain professionals shows that
manufacturing companies with clearer
visibility into operations and market
activity through supply chain analytics
can better foresee challenges and thus
respond to them proactively, increasing
both efficiency and profitability.
One of the biggest barriers to a wider
adoption of analytics in the marketplace is the lack of education and
misconceptions about the subject.
A lot of users, industry analysts and
consultants have not fully grasped the
difference between business intelligence (BI) and analytics. They continue
to consider simplistic query and reporting and OLAP drill-down capabilities
to be analytics, thus limiting themselves
to traditional BI systems that provide
simple alert, monitoring and dashboard
capabilities. While BI tools are very
important for answering questions such
as what, when and where an event
happened, they do not provide predictive insights that allow future business
decisions to be optimized.
And that is where true analytics come
into play. True analytical capabilities
such as forecasting, data mining,
predictive modeling and optimization
provide businesses with an understanding of why something is happening,
when it can occur again, and what will
35. SMBinsights
P34
Delivering analytics through a variety of channels
means that users in organizations of all sizes can
improve forecast accuracy, perform what-if analysis
and optimize resources – all without ever leaving
the comfort of familiar planning modules.
be the future impact of decisions,
so that outcomes can be optimized.
To succeed in the current economic
environment, businesses can no longer
rely on traditional BI tools that only give
you a view of the past. They must use
advanced analytics to get better
insights into the future to be proactive
rather than reactive.
Another popular misconception about
analytics is that only companies that
employ doctoral-level statisticians can
take advantage of advanced science.
The reality is that solutions today are
packaged in such a manner that even
novice and intermediate-level modelers
now can take advantage of advanced
modeling techniques via point-and-click
interfaces.
Earlier roadblocks of user resistance
and cost of integration with existing
technology are also no longer valid.
Advanced analytical capabilities are
now available via cost-effective channels such as software as a service
(SaaS), on demand, and from existing
ERP and SCM systems using serviceoriented architectures (SOA). Delivering
analytics through a variety of channels
means that users in organizations of all
sizes can improve forecast accuracy,
perform what-if analysis and optimize
resources – all without ever leaving the
comfort of familiar planning modules.
So what is stopping organizations?
Something that many of you can probably relate to: the sunk costs.
Companies have expended so many
resources in customizing and configuring existing SCM systems that they do
not want to commit any further resources to new, better technology – even if
the benefits over the existing system
are substantial.
ONLINE
Download the full report, Supply-Chain
Analytics: Beyond ERP SCM
sas.com/smb-beyonderp
Watch this webcast about adding analytics
to your supply chain:
sas.com/smb-supplychainwebcast
Before you resign yourself to the status
quo, ask what makes better sense in
the long run: continuing to sink more
money into maintaining an existing
system that is already behind the
times – or updating it with new, advanced technology that requires an
initial outlay but provides the robust
functionality required to survive in the
new economy?
Ritu Jain is the Global Marketing Manager for
Small and Midsize Business Solutions at SAS. She
is responsible for driving the company’s marketing
strategy in the SMB space. Prior to joining SAS, Jain
held a number of leadership positions in retail supply
chain management, driving strategic planning, global
sourcing, operations and production management.