Proving CX program value: How to tie scores to business results

Kelcey Curtis | Aug 1, 2017 Kelcey Curtis 08/01/17

Make customers happy. Get rewarded.

Even simple, tried-and-true strategies like that one are hard to execute consistently. That’s why leading brands are investing more resources into customer experience (CX) measurement programs, which allow them to collect, analyze, and act on customer feedback in real time. But with increased investment comes increased pressure to produce tangible results.

While an effective program will make more customers happy—and likely boost sales in the process—you still have to be able to connect those dots when budgets come under scrutiny. Otherwise your strategy might have too many gaps to justify keeping it in place.

To make a continuous case for program investment, you can fill in those gaps by determining how CX data correlates to business results at multiple levels. Here are some best practices to get you started:

Prove better customer experiences translate to location-level growth
No matter how big your business grows, it lives and breathes at the location level—which is why it’s important to understand how CX scores relate to (and impact) each location’s sales. One of the most common types of financial linkage techniques, correlation analysis ties measures like Overall Satisfaction, Likelihood to Return, and Likelihood to Recommend to location-level sales growth, which helps prove the direction and strength of the relationship between survey results and business results.

To go a step further and build a more sophisticated linkage model, you can supplement correlation analyses with multivariate regression analyses, which factor in additional data on location characteristics. The multivariate regression model pinpoints the amount of variance in sales growth accounted for by characteristics like:

  • Age of the location/recent remodels
  • Amount of nearby competitors
  • Regional unemployment rates

Data needed to make it happen:
For both correlation and multivariate regression analyses, it’s best to use a year’s worth of CX data and two years of financial or traffic data to ensure there’s a reliable number of responses per location. Factors needed for multivariate regression analyses typically fall into three categories: location descriptive information, periodic information, and outlying information.

Secure buy-in by showing the transaction-level impact of great service
While showing correlation to location-level sales is valuable, it’s not always the most compelling way to communicate program impact to different levels of the organization—and it may not be the right way to look at linkage for all industries. If you don’t have location-level linkage, the most direct way to quantify the financial impact of CX measurement is to match individual survey results to the amount spent by the respondent. Transaction-level analyses are the most detailed (and powerful) sources of financial linkage, as they help operations, franchisees, and other front-line employees understand the impact of their actions on each transaction.


Transaction analyses are particularly useful in industries like casual dining and specialty retail, where front-line employees have a more direct influence on customer spend. They enable you to dig deeper into the data to investigate specific areas of interest, including:

  • The impact of service standard execution by front-line employees
  • How ongoing promotions or limited-time offers (LTOs) affect sales
  • Proof that loyalty-building service behaviors boost average tips

Data needed to make it happen:
To conduct transaction-level analyses, you need to make sure there’s a unique identification number matching the survey to the transaction. Typically the survey entry code can accomplish this by including information like day/time of visit, location identifiers, and transaction numbers.

Engage the c-suite by demonstrating company-wide impact
Showing location-level sales growth and transaction-level financial linkage will help secure buy-in, but if you want the c-suite to be truly engaged, the best method is to prove the program’s impact on the entire organization over time—but that can take a while. Customers need to have an improved customer experience and then change their behavior (e.g., visit more often, recommend more frequently) as a result before sales reflect that change in behavior.

A time series analysis reveals how customer satisfaction and loyalty scores correlate with financial performance over time at the company level. This type of analysis is especially helpful when there’s a lag in the data that makes basic linkage techniques unclear. However, it’s worth noting that the length of the lag between improved service and improved sales varies depending on the brand's typical length of time between customer visits. For example, improvements for brands with long repurchase cycles like automotive or furniture retailers may not be as immediately obvious as improvements for QSR brands.  


Data needed to make it happen:
It may seem obvious, but time series analyses take time. To get sufficient sample, our best practice recommendation is four years of customer satisfaction/ loyalty data and five years of sales data, but the analysis can be performed with a minimum of 30 months of customer satisfaction/loyalty data and 42 months of sales data.

Making the case for your program
An effective customer experience program should help you create great customer experiences. But for the program to last, it takes more than smiling customers—ideally data validation can be demonstrated through connections to tangible outcomes. There are many ways to do that, but the most convincing is to link it to your bottom line.

To learn about financial linkage best practices in more detail, download our white paper.

Kelcey Curtis, MA
Research Manager

Customer Experience Update