There are 3,874 vendors listed in the 2016 Marketing Technology Landscape, and the phrase “MarTech stack” yields over 50,000 Google results. What’s a rational way to decide what you actually need?
Join experts from DemandGen and Openprise as they provide a strategic framework for deciding what systems and what data you need to be successful.
17. • Establish a baseline
• Input to the maturity model /
roadmap
• Know areas of weakness
• Know where low-hanging fruits are
• Monitor quality metrics
Start With Data Quality Assessment
18. Clean, Accurate, & Updated Data != Quality Data
Do you have the RIGHT
leads?
Do you have USEFUL
segmentation?
19. • Profile based scoring
• Buyer persona
• Account based marketing
• Personalization
• Campaign execution & ROI analysis
Segmentation Starts With Why?
20. Common B2B Dimensions Custom Dimensions
• Job function / sub-function
• Job level
• Buyer persona
• Industry
• Company size
• Region / territory / metro
• Language
• Number of job openings
• Size of vehicle fleet
• Number of mobile devices
• Number of stores
• Competitive products
• Complimentary products
• Channel
Which Dimensions & How Many Segments?
22. • Difficult to correct false positives
and overlaps
• Even more difficult to identify
false negatives
• Can’t analyze and report on
segmentation results
• Can’t be used by other systems
But I’m Already Doing Segmentation With Filters
The Problem With Filters You Can’t Do Analysis Like This
Add Segmentation As
Part of Your Data!
23. What You Will Need To Do Segmentation Right
Segmentation Mapping Data
A Data Automation Platform
25. This even assumes data from the different sources are clean.
CA
vs.
California
Normalize data standards
+1 415-555-1212
vs.
(415) 555-1212
Normalize data format
Toyota Motors U.S.A.
vs.
Toyota Motor Sales USA
Correlate non-exact data values
Normalize segmentation
11-50 employees
vs.
20-100 employees
Unification Challenges
26. Unification Tips
You don’t have to have a
central data warehouse
Leverage the systems you
already have
Avoid a master data schema
at all cost
Translate and map what you
need as required
There is no “perfect” or even
“best” data vendor
Know which data parts you
want from which vendor(s)
Have a unification strategy
before buying any data
Digest data immediately
after acquisition
Data unification is not easy, but the companies that can make it happen has an absolute unfair advantage in terms of gaining new insights, creating better engagements, and execute with high efficiency and speed.
Here are 4 tips I have for marketers who are embarking on data unification projects:
The knee jerk reaction when people talk about data unification is you need a central data warehouse to pull all the data in one place. You actually don’t have to. Depending on the scope and targeted users for the project, you can very will use your CRM or marketing automaton platform as the system of record for the reconciled data.
The second knee jerk reaction is that you need a master data schema that can unify all the data you have from difference sources. It is extremely difficult to come up with such a schema that can be maintain with reasonable effort. This is probably the number one reason why master data management projects fail and they do fail more often than they succeed. It is just too rigid. Instead, develop the ability to map and translate whatever subset of data you need for each applications as you need it.
There is no perfect data set that will fill all your needs. As your data management program gets more mature, you will most likely come to the conclusion that you need multiple data vendors to assemble the data you need. Get to know which data vendor excels at what type of data and which data parts you need from each data provider. It’s up to you to assemble the data that is perfect for you.
As we mentioned when introducing the data management framework, we recommend you do not buy any data until you have a unification strategy on how you will digest new data. If you don’t have a unification strategy, you won’t be able to digest the acquired data immediately, and if you do not digest it immediately, it will accumulate and lead to the data hording problem.