Good morning. I am going to start today’s session with a quick poll. I want to get a sense of the different roles we have in this session:
Who is in marketing? Sales? Products?
(based on the audience segments focus story to meet role expectations but emphasize this is largely focused on marketing)
I am very pleased to be here with you today to share with what I hope will help you understand the value of applying analytics to improve your marketing effectiveness. I will be joined by my colleagues, Doug Smith, Manager of the Marketing Analytics center, and Bob Slaker, Program director for direct response marketing within our Predictive Analytics business unit. We will start the session by sharing IBM;s point of view on what the marketing profession is becoming and the requirements to be successful. I will then invite my colleagues to share some insight about IBM’s journey to become an analytics driven marketing organization. We will spend the last portion of our 45 minutes reviewing a slice of IBM’s own story, not the whole story, but rather a portion which wil help you understand how to develop a digital thumb print of your customers.
Marketing organizations have long been tasked in 3 key areas: Knowing who customers are; What they need to market and how; and finally stewarding the brand.
We now need to understand customers as individuals (I bought shoes for my wife once online… that first interaction was a great experience…However…the online retailer has now begun to market me other womens items…with womens messaging. Do they really know me? )
we need to ensure the processes are there to engage customers in an omnichannel world (Think about us as we sit here or our living rooms….we have an ipad, iPhone, laptop, television, face to face interactions…..as marketers we need to figure out how to take advantage of all these channels in a coordinated way..)
and we also need to ensure that our brand and culture reflect each other. (Our CEO talks extensively about brand and culture. If the two were to have a battle which one wins out? Thoughts? I submit culture wins…in fact, it would eat the brand for lunch….)
While the demands on us marketers are changing, many of the questions remain the same. We still need to understand who is responding, which customers are our best customers, which offers are resonating…but understanding what has happened or what is happening is simply not enough…. Looking at what has happened and what is going on now is like doing an autopsy on a dead guy. Sure, we can answer what killed the patient but wouldn’t it have been better to be able to predict what was going to do this guy in BEFORE it happened?
As marketers, we need to get to the point where we can understand buyers as individuals and predict what they are likely to do. In a world where the majority of our marketing activities are face to face we can see our buyers’ reactions …You can see the reaction on these athletes’ faces…dejection..As a marketer, now is not the point where I want to sell these guys insurance for their gold medals (unless you are Canadian)….We can see it, it is palpable…But what about this?
Here is a graph of marketing inquiries for a campaign…what does that tell us about our buyers’ attitudes, feelings, emotional state or behavior?? It can tell you a lot… with all of the online activities today it is possible to capture these interactions…In fact, we as consumers are leaving our fingerprints all over the place for enterprising folks like use to find! As marketers we have got to get to the point where we can identify buyers as individuals…to a point where we can pick out our buyers’ “digital thumb prints” among the thousands of tire kickers to put the right offer in front of those buyers who we can convert to a Sales Accepted Lead…but how do we do that?
Well, we need to approach our interactions differently…holistically if you will. We need to treat every interaction as a line or dimple in the thumb print … There is a wealth of data out there (speak to the 4 areas and what can be captured) We marketers must capture it, analyze it and act on it…
Poll audience: Who here analyzing their waterfall performance? Inquiry to MQL to SAL to WR? Who here is analyzing what lead to that success? Now, who here is incorporating survey data or social media data into that same funnel analysis to get the whole picture and predict success or failure? (Depending on answers…react surprised or not.) We at IBM are on a similar journey…we are looking for ways to indentify those buyers’ thumb prints and offer them the right thing at the right place at the right time…I would like to invite my colleague Doug Smith to share a few of the things we are doing to
So first off, Brendan is introducing me to share IBM’s story to try to explain our journey over the past decade or so.
So first off, Brendan is introducing me to share IBM’s story to try to explain our journey over the past decade or so.
Brendan has set the stage. Doug has given history and how IBM goes about analytics.
Story about a company … which happens to be a one that IBM acquired.
SPSS, a 40 year old company that made and sold statistical and data mining products … predictive analytics.
Academic pedigree … sold primarily to analysts, statisticians, mathematicians … people with PhDs, as well as students learning to hate statistics.
Who in the room used SPSS in college?
When I arrived at SPSS in 2001, the company was working aggressively to grow beyond our traditional audience.
I looked up the numbers the other day … there are about 4,100 colleges and universities in the US. On the other hand, there are 5.9 million employer firms in the US … with 655K with 20 or more employees, and 129K of those with 100+ employees.
Our challenge was that predictive analytics was not considered “mission critical” by many companies … until they started getting buried by web data.
You’ve probably heard of “big data”
We were solidly imbedded in traditional audience … academics, statisticians, financial analysts … people with advanced degrees and statistical backgrounds.
But to grow more rapidly, we needed to move beyond our traditional audience …
Small orgs … buying software tools
Departments in larger orgs … buying tools and marketecture
Enterprise users … tools, marketecture and solutions
Also, while we were technically a B2B company, much or our approach was more like B2C … or as we called it, Business to End User. Much of the process was around selling to an individual who had a specific need for their job or department, and could make purchases with limited approval required.
Our marketing process was complex in that it involved a variety of prospect and customer types, all of which were spinning off lots of data.
The good news, is that we had amazing technology at our fingertips, however …
While we were creating and selling analytics software …
We operated as a marketing and sales org, so our use was more ad hoc.
In some ways, we could have been selling pots and pans.
(we aren't the only company out there like that ...)
Our experts were out helping other organizations to take advantage of predictive analytics … but that left us marketers to our own devices ...
Great tools
… but not the full stack
We recognized the need to leverage our own analytic capabilities on our own data, and started down the road to building out end-to-end automation, tracking and analytics.
About that time, we got bought by IBM.
We were acquired by IBM in 2009 … this gave us the opportunity to take advantage of being a “customer”
Gaps in expertise were filled, access to systems made available. We were enabled to step back and look at our business in a new way.
Much to our delight, we were now connected up with organizations within IBM who had been using our technologies to conduct business exactly as we had always advertised … but had not always been able to do ourselves. Ironic, isn’t it?
Immediately we dove into our own business, applying analytics.
Who in the room is familiar with the Sirius Decisions Demand Waterfall?
One of the first places we went was the demand waterfall.
We began the process of ... examining each transition point in waterfall.
Identifying points of interaction
Between people
Between campaigns
Between media
Between timing
Studied the predictive factors that influenced conversion at each step, and the routes customers and prospects took to become wins
What were the factors?
Demographic, firmographic, psychographic, sales approach.
Then we looked at the interactions between the buying cycle, marketing cycle and sales cycle
We found that too often we were looking out instead of in.
Marketing was looking at their process to bring in customers …
Marketing - Create Awareness, Educate, Convince
Sales was looking at their process to close people
Sales Qualification Cycle - Budget, Authority, Need, Timing (or NTAB)
And, all the while, oblivious to how marketing or sales looked at them, the customers and prospects were going through their own process.
Consumer’s Buying - Awareness, Interest, Consideration, Evaluation, Purchase
Then when one considers the interactions between each of the stages … a very complex picture emerges and it all needs to be framed in the context of customer and prospect behaviors … where they are leaving their thumbprings.
So, how’d we do?
First … “consistently capture and analyze interactions developing a digital thumb print.”
We have identified the interaction points, in search, our web site, content syndication, events etc. and are capturing explicit behavior.
We've learned that by the time we see explicit behavior, customers and prospects are often further down the buying cycle … sometimes ready to place an order ... so need better early warning to make sure the order is placed with us
We need to become better at identifying and capturing implicit behavior (such as unregistered site visits). That initiative is currently in development to get that better early warning in place.
“Predict client behavior to segment customers and optimize campaign ROI”
We are using analytics to model customer and prospect segments to identify most likely to respond, most likely to become opportunity, most likely to win
From a marketing campaign investment viewpoint, we’re always looking for ways to improve the “odds” of a success. We've learned that while leveraging larger segments available in prospect data does improve our odds of campaign success, defining smaller, more defined segments within customer behavior has maximum impact. However, this targeting has to be balanced with the volume available … and your relationship with sales.
With more careful targeting and segmenting, you’ll likely see a fall off in the number of leads going through to sales. It requires a bit of a culture shift to move from “lots of marketing responses” to a much smaller number of qualified leads.
We are still under pressure to "produce more leads" at various times. Our team has had to provide more education into our sales partners about the quality of lead that can be provided, in smaller numbers.
“Synchronize marketing processes to create a closed loop and global view of the customer.”
We are tracking, capturing and centralizing our data to identify the lifecycle and path of customers through our interactions. This understanding of the customer has enabled us to field the right campaign to a given audience at the right time
We've learned that while we can use that global view to field the right campaign, we need to get better at mapping that global view to agreed upon campaign activity … still have demand for "renegade" campaigns from sales that run through the audience with less than a strategic approach to that customer.
“Embed predictive capabilities to drive personalized campaigns and micro-targeting.”
We are producing personalized campaigns at a gross level, and have the capability of micro-targeting, but are sometimes resource-challenged to implement.
We've learned that having very detailed information can lead to "paralysis by analysis" as one sorts out and prioritize the opportunities to match with resources. Our approach has been to focus in on key high performing segments before expanding out into the realm micro-targeting.
We have been leveraging our integrated marketing and sales automation systems to begin the process of building ever more personalized campaigns … but we have to continue developing leads for sales while we carry that out.
“Gain real-time intelligence and cross-channel reporting and benchmark capabilities of marketing campaigns.”
We recently made significant configuration changes in our internal systems to accelerate our ability to see what is happening, and act upon it. Campaign responses from nearly all our channels flow into the same system, and are managed in the same way. This has allowed us to identify the relative success of campaigns, and combinations of campaigns.
We've learned that we have had to make a real shift in view of understanding from the "how did the individual campaign perform" to "how are we managing this audience segment“ as we progress to “how do we reach this individual appropriately.”
In the past, a lead equaled a response. Now, a lead is a respondent … with perhaps multiple responses and touch points.
Notice that Albert’s formula is “Analytics = Success squared” … it enables to you move from linear growth to geometric growth.
“Combine attitudinal and survey-based data with social media sentiment to anticipate and target new segments.”
While we have historically been expert at surveys and the use of their data, like many, our use of social media sentiment internally in the SPSS business is in its infancy. However, we are beginning the application of our own Social Media Analytics technology, and later this year I expect we will have some interesting things to talk about here.
One key thing we have learned is this sort of data has to be carefully calibrated with actual behavior ... as people frequently will say one thing and do another. In some cases, it is better to ignore the apparent meaning of the "text" related with such data and just code it, and test that coding against behavior.
Give up old views of how campaigns are measured
Become better advocates/collaborators with our sales partners
Recognize that it takes more than great creative to get results
Understand that we no longer control the message
Manage the fact that our "radar" has to go out further, as the first time prospects emerge in the traditional view, they are very far into their buying cycle
Recognize that customer and prospect organizational behavior affects individual behavior
As Brendan and Doug laid out, there is a set of items that can be applied to build out and leverage the understanding of customers.
So, we set about applying and maximizing the full set of analytic capabilities to drive our business.