One key predictor of future company growth is customer loyalty.

It’s such an important customer metric, I have a chapter dedicated to measuring loyalty in Customer Analytics For Dummies.

There is evidence that customer loyalty predicts future growth better than customer satisfaction. That is, customers can be satisfied with a product and yet still choose not to purchase it again.

Customer loyalty is best measured using a mix of attitudinal and behavioral metrics, and a repurchase rate is a good one to start with.

When customers have a choice in products or services and make repeat purchases (airline tickets, consumer electronics, rental cars, hotels, software licenses), the percent of customers that purchase again from the same company provide a good indication of customers’ loyalty to that company.

Building a Repurchase Matrix

A good way to visualize the repurchase rate, and at the same time discover where customers are defecting to, is a repurchase matrix. I learned this technique from Fred Reichheld, where he explains the repurchase matrix in his seminal book The Loyalty Effect. Repurchasing can happen frequently, say weekly for business travelers buying airplane tickets, or over a long time, say over a decade, as is often the case with businesses upgrading enterprise software.

A repurchase matrix takes the repurchase rate and displays the relative rates for a set of competitors or comparable products within the same company. To build one, compute the percentage of customers that purchased a product from one company and then purchased from the same company again (repurchase) or defected to purchase a competitor’s product.

For example, the table below is a repurchase matrix for laptop computers using hypothetical data. The first cell indicates that 68% of customers that bought a Lenovo computer ended up buying a second one (they repurchased). The second cell indicates that 3% of the customers that first bought a Lenovo laptop chose Dell the second time around.

First Purchase Repurchase
  Lenovo Dell HP Apple Asus
Lenovo 68 3 4 2 3
Dell 14 43 4 2 5
HP 13 5 42 1 5
Apple 4 1 1 77 7
Asus 3 1 1 8 38

Table 1: Numbers are percentage of customers that repurchased the same brand of laptop (hypothetical data).

A few things to note about repurchase matrices:

  • The values on the diagonal that start in the upper left and go to the lower right are the repurchase rates for the same product. Scanning down allows you to see the repurchase rates for all brands (e.g. 68% for Lenovo, 43% for Dell, 42% for HP etc).
  • The values off of each diagonal are the competitors to which the customers are defecting. For example, 13% of HP customers are defecting to Lenovo.
  • The values often don’t add up to 100%, because not all competitors are considered or customers don’t purchase again in the product category.

In looking at this example data, you can see that Apple customers are about twice as loyal as Asus’ customers. Apple has a repurchase rate of 77% compared to 38% for Asus.

Collecting Repurchase Data

There are three common sources for building repurchase matrices: historical customer data, internal surveys and third party research.

  • Historical Customer Data: Repurchase data lives in most company’s sales automation or inventory software associated with customer accounts. Pick a common repurchase period (e.g two years for laptops) to look for evidence of repurchasing. If you have a product that isn’t purchased frequently, it could take years to build a repurchase matrix with actual data. To speed up the process and gauge your customers’ loyalty, you can ask customers’ intent to repurchase.
  • Survey (Intent): In a survey, ask your existing customers who they purchased from before and who they intend to purchase from in the future. It’s easy to add these two questions to an existing customer satisfaction survey and it provides a quick gauge of repurchase and defection rates by competitor. But remember that people don’t always do what they say they’re going to do. Be sure you and other product stakeholders know the difference between the attitude (intent to repurchase) and behavior (actually repurchased) when interpreting the data.
  • Third Party Research: A final source is from third party research firms (including ours). While research firms are unlikely to have access to historical customer data from multiple companies, they can easily survey customers on their current, prior and potentially future brands for any product or service category.



Company growth depends on customers not only making that first purchase but also making repeat purchases. A repurchase matrix gives you a good idea how many of your customers make a second purchase with you or with another company. Ideally you can use actual repurchases from historical data; otherwise use the intended repurchase rate from customer surveys or third party research as a substitute.