Marketers dream of a situation in which they could capture the information from every form consumers ever filled out, every transaction they completed, or every exchange they had, providing such a clear picture of consumers’ needs that an invaluable, highly relevant, personally accurate relationship can be created.
Granted, this perfect amount of behavioral content may not be readily available, but there remains a great and growing volume of information, which, if used correctly, can help marketers create relationships in which they can predict consumer purchases before that person has even decided to buy.
Marketing analytics is the key to extracting the value from this information, and enabling this next phase of marketing, with enticing rewards. Like increasing response rates, revenue per customer, product holding and loyalty. And reducing cost of acquisition, marketing-spend, wastage, and lapse and churn.
Three routes, one destination
There is no wrong or right way to learning who your customers are. There is an iterative cycle of learning through which you improve your knowledge. The more accurate your knowledge, the more accurate your targeting.
For these reasons, more and more organizations deploy a range of marketing analytics every time they embark upon a project. These help to understand and explain customers and their behavior, gain the ability to predict their behavior and buying patterns, and to make our customer targeting and messaging more accurate.
They fall into three primary areas: factual, descriptive and predictive analytics. Each is valuable, but combined they can elevate your customer insight to exciting levels.
ò A clean, de-duplicated single view of the customer
ò Make the facts usable (i.e. “Date of birth” turned into “Age”)
ò Create new facts from existing facts
(e.g. recency, frequency, preferred channel)
ò Missing or unclean data
ò Duplicated data
ò Convoluted facts with no real meaning
ò Facts with no applicability to your business or customers
ò Use of indexing and penetration indices
ò Simple cross tabs to indicate populations
ò Geographical mapping
ò Comparative analysis
ò Vertical specific e.g. market basket analysis
ò Clustering models
ò Combination of above for segmentation
ò Build predictive models for:
ò Buy a specific product
ò Buy a specific product combination
ò Behave a certain way (i.e., churn)
ò Predicting lifestyle variables e.g., income, age, number of children
ò Complex combinations: e.g., best next action, scorecards, etc.
Bill Marjot was the chief marketing officer at SmartFocus, Bristol, Britain.