Skip to content
Back to Blog bigbetterdata-1
Finance calendar    Jun 02, 2018

Applying, Not Bigger, but Better Data to Customer Analytics

How customer-centricity, CLTV and better customer data drives prioritization, profitability and greater shareholder value. Key takeaways from Edison's 2018 CEO Summit.

At Edison, we collect a lot of data from our portfolio companies: through quarterly financial reporting, our annual Growth Index, or that new data point we urgently need right now, we are constantly digging in an extra layer deeper – likely more so than the rest of the growth equity industry. We find adding that additional dimension and understanding how it compares across the other current and exited Edison companies paints a clearer picture of performance. With our recent Growth Index, specifically, we looked at 40 active companies. Limited data set? No way. As Wharton Professor Eric Bradlow put it at our Edison CEO Summit earlier this month, rather than adding more rows, we focus on adding better columns. So, what are the better columns and what can we help our companies do with this data?

CLTV is King

When you have an unprofitable product line, should you shut it down? What if your top five customers use the product and the other users are all negative-margin customers? It’s not the product line that we need to get rid of. Segmenting value by customer (or CLTV) is a key step in understanding how to prioritize in order to unlock growth in your business.

We all know the importance of CLTV and the role it plays in driving long-term shareholder value. Does that mean we should devote all of the company’s time and attention to the highest CLTV customers? Not quite. There’s a reason these customers are driving the greatest level of CLTV – they’re loyal customers! They’re not going anywhere. So, who do we devote our attention to?

How to Rank Your Customers

Last year, I certainly thought HBS professors were wicked smart, but Professor Bradlow blew my mind with this one: rank your customers using the derivative of CLTV. I owe my 11th grade math teacher an apology; we can actually use calculus in the real world! Well, the SaaS world, at least.

It’s all too common for growth companies to apply the best/most resources to highest paying customers when we should be prioritizing the customers for which one of the following will increase:

  1. Frequency of use
  2. Basket size
  3. Probability of retention

Unless you can improve one of these three metrics, there’s no point investing time in a customer. Of course, there was a CEO in the crowd who noted his company’s model didn’t really allow for improvement in any of the above metrics. Now what do we do?

Apply Current Customer Data to Prospects

That CEO’s question was legitimate – his company has no influence over frequency of use, they didn’t have additional modules to upsell, and gross retention is already over 97%. However, they have a sizable addressable market and robust pipeline. Having a ton of data on the company’s hundreds of customers, this CEO has a deep understanding of his ideal customer profile characteristics, and now knows to prioritize prospects with those characteristics. But what do we do with those negative-margin customers we talked about earlier?

How to Fire Customers

Wouldn’t it be great if we could just call up our negative-margin customers and tell them, “I’m actually losing money on our deal; do you think we can change that?” Well, you can do that without even making that ridiculous call – try one of the following:

  1. Raise prices for that customer segment
  2. Reduce resources (support, customer success, professional services) dedicated to the customer segment

One of two things will happen – the customer might churn, and now you’ve rid yourself of an unprofitable customer, OR perhaps demand was less elastic than we initially thought and the customer is willing to pay more: now they’re profitable!

Technology and what you can measure about your customers have changed the face of Marketing. Not all companies will have all of the necessary data at their fingertips – you’ll have to go find the best way to collect it. Once you do, these unprofitable customers are often the best segment on which to run experiments – the newly acquired data might even be more valuable than their revenue.

(Btw, Professor Bradlow leads the Customer Analytics Initiative at Wharton. His team welcomes corporate participation as they work to enrich their data and methodologies across industry sectors. Participation is free.)