Projection: How to map results from a sample onto your customer base

There are two aspects of deep analytics: projection and predictive modeling.  In the video below, we are focusing on projection:

So, what is projection?  Projection is taking valuable information from a sample of customers and mapping those results onto the rest of your population.

There are two types of variables that you need to start:  customer descriptors and customer experiences.  Customer descriptors can be age, income, activities, geography, tenure, vertical, size of organization, etc.  For experiences, you look at interactions or touch points with your customers – how many calls they made to customer service, how many times sales interacted with them, etc.

How is this valuable to you?  Three reasons.

The first is using projection to identify customers with strong relationships and weak relationships with your organization.  For example, your results show that loyal customers give you, on average, 63% of their spend.  When you project the variables onto your population, you will find that there are some loyal customers that give you more than 63%, and some that give you less.  These customers love you and your product, so your first goal should be to find out why you aren’t getting more share from them.

On the other hand, vulnerable customers that give you a high share of spend are a threat to your organization.  With projection, you can now identify and proactively fix problems and manage these accounts.

Another way to use projection is to look at your prospects.  Which prospects are more likely to be loyal and which are likely to be vulnerable?  If you are in any way concerned about Customer Lifetime Value, you really want to focus on winning those prospects more likely to be loyal and avoid those who are likely to be vulnerable.  Loyal customers are easier to retain and have a higher likelihood of spending more with you over time.

Finally, in looking at prospects using the above strategy, we avoided potential customers because they were likely to be vulnerable.  Using projection, you can find out the reasons these customers would be vulnerable – poor local branch performance, weak sales reps, etc. and fix them.  You don’t want to win prospects if you’re likely to fail them; you need to fix these issues before you bring them in.

In summary, projection is a great tool to understand how to grow your company more effectively.  To to this, you target very specific strategies at the right customers.  To learn more about projection and how you can use it in your own organization, email us at

Posted in Blog, Deep Analytics, Insights, LRC Blog, Videos and tagged , , , .