Pitting Big Data Versus NPS Isn’t The Way To Success In CX Measurement
You might have seen a recent blog by Tony Cosentino on how “Big Data Analytics Will Displace Net Promoter Score (NPS) for Measuring Customer Experience” because “NPS is prone to error, lacks a causal link with financial metrics, and lacks actionable data.” And while Mr. Cosentino’s blog highlights a critical issue in CX measurement, it only tells part of the story.
The problem in most CX measurement programs — whether that company uses NPS or not! — is that CX pros rely too much on surveys. That’s because of a number of factors:
- Inherent survey bias. People in CX roles come from a market research background, so surveys are a natural competence and a go-to strategy for them. Surveys are a good tool to quantify customers’ perceptions. But we know that surveys capture feedback for too few interactions from too few customers too late in the game. However, some companies, including JetBlue, are going beyond surveys in their CX measurement. They use data like location data, social media insights, and new tools like predictive analytics to supplement and replace surveys. Where those companies go, you should follow. Read more about it in my report “Beyond Surveys: New Tools For More Effective CX Measurement.”
- Needle-in-the-data-haystack problem. Even CX pros who want to use behavioral data more struggle to know which behavioral or operational data they should even be looking at as a leading indicator of customer sentiment and loyalty. That's because they try to attack this topic inside out (let's put all the data we have in a bucket, shake it, and let the insights tumble out). That doesn’t work. Instead, CX pros need to start outside in (i.e., with the customer journey and make their way down from expectations of customers to possible metrics they should look at). Read more about how to use this approach to generate a better hypothesis about the link between behaviors and CX in my report “How Journey Maps Improve CX Measurement Efforts.”
- Data silo challenge. Lots of CX pros I speak with find it hard to get at data about revealed behavior. They are frustrated that data is in disparate sources, and they find it difficult to get to insights at a customer level (as opposed to a segment level), which is where they need to have the data to do these kinds of predictive analytics.
But even though I agree that NPS isn’t the ultimate metric for each company (but then, let’s be pragmatic, which metric is . . . ), Mr. Cosentino’s blog tells only part of the story. Firms need (big) data and customer feedback because:
-
Humans need a concept to rally around. The attention span of humans and the culture of organizations are such that both need a concept or a rallying cry around which they can align. Over time, we have seen several of those come and go — the zero-defects quality movement or beacon metrics like customer satisfaction, NPS, and customer effort.
-
Revealed behavior doesn't help with the why. Surveys — and qualitative research, for that matter — do help. That’s why many companies rely on them. Take, for example, an employee from DropBox who mentioned at a recent conference that the company has lots of behavioral data about its users but realized it also needed to understand why customers act that way. So it beefed up its qualitative research capabilities.
-
Actionability needs more than data. We know that successful companies don't have NPS or customer effort. They have a system around a beacon metric in which they identify what drives those metrics. And the very successful companies are able to translate those drivers into actual behaviors for employees throughout the organization. You can read more in Sam Stern’s case study “How Hampton Hotels Built And Sustains Its Customer-Obsessed Culture” and the Pitney Bowes example in my report on “How Three B2B Firms Measure Customer Experience.” And even companies that invested in big data practices and technologies still “paddle in a sea of data without the ability to prioritize results and find insights they can use,” as my colleagues Ted Schadler and Brian Hopkins found.
So far, I haven't seen other approaches be as successful as a beacon metric for rallying the organization around CX. If you have, I'd love to hear about it! And if you have more examples of companies that use (big) data and analytics to supplement or replace CX surveys in their CX measurement program, please let me know.