Customer Loyalty Research: What the Data Tells You

Customer loyalty research tells you what customers say they value, how they describe their behaviour, and which brands they claim to prefer. What it rarely tells you, without careful interpretation, is why they actually stay, why they leave, or what would change either outcome. That gap between reported loyalty and real loyalty is where most retention strategies quietly fall apart.

Used well, loyalty research is one of the most commercially valuable inputs a marketing team can have. Used poorly, it produces confident-sounding slides built on shaky foundations, and organisations make expensive decisions based on data that was never as solid as it looked.

What Does Customer Loyalty Research Actually Measure?

This is the question most teams skip. They commission or consume loyalty research and move straight to the findings, without pausing to ask what the instrument was actually designed to capture.

Most loyalty research sits in one of two camps. Attitudinal research asks customers how they feel: how likely they are to recommend, how satisfied they are, how committed they feel to a brand. Behavioural research looks at what customers actually do: purchase frequency, basket size, churn rates, category share. Both are useful. Neither is complete on its own.

I spent a period judging the Effie Awards, which meant reviewing a significant volume of effectiveness cases from across the industry. One pattern appeared repeatedly: brands presenting strong NPS or satisfaction scores as evidence of loyalty, when the underlying commercial data told a more complicated story. Customers said they loved the brand. They were also buying from competitors at the same rate as before. Stated loyalty and enacted loyalty are different things, and conflating them is an easy mistake to make when you are close to the data.

The distinction matters because it changes what you do next. If your research is primarily attitudinal, you know how customers feel. If it is primarily behavioural, you know what they do. You need both to understand the relationship between perception and commercial outcome, and to identify where the gaps are.

Why Loyalty Varies So Much by Industry

One of the more useful things loyalty research confirms is that loyalty is not a universal construct. The dynamics that drive retention in a subscription software business are structurally different from those in a grocery retail context, which are different again from financial services or a local independent business.

Some of this variation comes down to switching costs. In categories where changing provider is genuinely painful, whether because of contract terms, data migration, habit, or perceived risk, customers stay even when satisfaction is moderate. That is not loyalty in any meaningful sense. It is inertia. Measuring it as loyalty flatters the brand and obscures the real risk: the moment switching becomes easier, those customers are gone.

Research from MarketingProfs has documented how consumer loyalty and satisfaction vary considerably by industry, with some sectors showing strong correlation between the two and others showing almost none. That variation is not a footnote. It is the core finding, because it means you cannot apply a generic loyalty framework and expect it to work. You have to understand the specific category dynamics your customers are operating within.

When I was running an agency and we were working across thirty-odd industries simultaneously, this was one of the things that consistently surprised clients. They would come in with a loyalty programme or a retention strategy borrowed from a different sector and wonder why it was not producing the same results. The answer was almost always the same: the category dynamics were different, the switching costs were different, and the reasons customers stayed or left were different. Research that did not account for that was not useful research.

If you want to understand the broader mechanics of keeping customers, the customer retention hub covers the strategic and tactical landscape in depth, from measurement to programme design to the commercial case for retention investment.

The Problem With How Most Loyalty Research Is Conducted

I do not reject research. I take it seriously. But I have seen enough of it commissioned, presented, and acted upon to know that the methodology is where most of the risk lives, and it is also where most teams spend the least time.

Survey-based loyalty research has a structural problem: it asks people to reflect on and articulate feelings and intentions that are often not consciously formed. When someone answers a question about how likely they are to remain a customer, they are constructing an answer in the moment, influenced by how the question is framed, what has happened recently, and what feels like a socially acceptable response. That is not a fatal flaw, but it is a significant limitation that should inform how much weight you put on the findings.

Sample composition matters enormously. If your loyalty survey is sent only to customers who have made a recent purchase, you are already talking to a self-selected group. Customers who have drifted or churned are not in the sample. The result is research that tells you how your most engaged customers feel, presented as a picture of customer loyalty overall. That is a meaningful distortion.

Timing matters too. Loyalty sentiment captured immediately after a positive service interaction will look different from loyalty sentiment captured three months later when nothing notable has happened. Neither is wrong, but they are measuring different things, and treating them as equivalent produces noise rather than insight.

The question I always ask when someone presents loyalty research is: who is not in this data, and why? That question surfaces the gaps faster than almost anything else.

What Behavioural Data Adds to the Picture

Attitudinal research tells you what customers think and feel. Behavioural data tells you what they actually do. The most useful loyalty research programmes combine both, because the gaps between them are often where the real commercial insight lives.

Purchase frequency, recency, and category share are the core behavioural signals. A customer who buys from you regularly but also buys from three competitors is not loyal in any commercially meaningful sense, even if they score you highly on a satisfaction survey. A customer who has not purchased in six months but still opens your emails and engages with your content is in a different position again, not yet churned, but drifting.

Propensity modelling takes this further. Rather than describing loyalty as it currently exists, it uses behavioural signals to predict which customers are at risk of leaving and which are candidates for deeper engagement. Forrester’s work on propensity modelling for account risk and upsell is a useful reference point here, particularly for B2B contexts where the commercial stakes per account are higher and the signals are more complex to interpret.

The practical implication is that loyalty research should not be a standalone survey exercise. It should be part of a continuous data programme that connects attitudinal signals to behavioural outcomes, tracks changes over time, and feeds directly into commercial decision-making. That is a higher bar than most organisations currently meet, but it is the bar that separates research that informs strategy from research that fills a slide deck.

Satisfaction Is Not the Same as Loyalty

This distinction gets blurred constantly, and it costs organisations real money.

Satisfaction measures how well an experience met expectations. Loyalty measures the likelihood of continued, preferential behaviour over time. They are related, but the relationship is not linear and it is not guaranteed. You can have highly satisfied customers who leave the moment a competitor offers a better price. You can have moderately satisfied customers who stay for years because switching feels too complicated.

The practical implication is that optimising for satisfaction scores is not the same as optimising for retention. Both matter, but they require different interventions and different measurements. A team that treats CSAT or NPS as a proxy for loyalty is working with an incomplete model, and the decisions that flow from that model will be correspondingly incomplete.

I have sat in enough client reviews where a rising NPS score was presented as evidence that retention was improving, only for the churn data to tell a different story. The two metrics were moving independently. The satisfaction improvement was real, but it was not translating into changed behaviour. Understanding why that gap exists, and what would close it, is a more commercially useful question than celebrating the satisfaction score.

Content plays a supporting role in this too. Unbounce has written about how content underpins customer retention, and the argument is sound: consistent, useful content reinforces the relationship between purchase moments and keeps a brand present and relevant in a customer’s consideration. But content does not create loyalty on its own. It supports the conditions in which loyalty can develop, which is a meaningful but limited role.

How to Read Loyalty Research Without Being Misled by It

There is a version of engaging with loyalty research that is essentially passive: you receive the findings, you note the headline numbers, and you use them to justify decisions you were already inclined to make. That is not analysis. It is confirmation.

The more useful approach is adversarial in a constructive sense. You read the research looking for the things it cannot tell you as much as the things it can. You ask about sample composition, question framing, timing, and the gap between what was asked and what you actually need to know. You look for the findings that are inconvenient or surprising, because those are usually where the real insight is.

A few specific questions worth applying to any loyalty research you encounter:

Is the difference between groups statistically meaningful, or just numerically different? A two-point difference in loyalty scores between two customer segments might look significant on a chart and be completely within the margin of error. If the research does not address statistical significance, treat the granular comparisons with caution.

Are the findings directionally consistent with behavioural data? If the research says loyalty is improving but churn is flat or rising, one of those things is wrong, or they are measuring different things. Either way, you need to understand the discrepancy before acting on either.

What was the response rate, and who did not respond? Low response rates introduce self-selection bias. Customers who feel strongly, either positively or negatively, are more likely to complete a survey than customers who feel indifferent. Indifference is often the most commercially important segment, because those are the customers most likely to drift without actively deciding to leave.

A/B testing frameworks can also sharpen your understanding of what actually drives loyalty behaviour, rather than what customers say drives it. Optimizely’s work on using A/B testing to increase customer retention is a practical illustration of how behavioural experimentation can complement attitudinal research, giving you a more grounded picture of cause and effect.

The Commercial Case for Taking Loyalty Research Seriously

None of this methodological rigour matters if the organisation does not treat loyalty as a commercial priority in the first place. And in my experience, that is where the real problem often sits.

Retention gets less budget, less attention, and less senior focus than acquisition in most organisations. The reasons are partly structural, partly cultural. Acquisition produces visible, attributable activity. Retention is more diffuse, harder to measure cleanly, and less exciting to present. The result is that loyalty research, even when it is well-conducted, often sits in a document that nobody acts on.

When I was turning around a loss-making agency, one of the first things I looked at was the client retention rate and what was driving it. The data was uncomfortable. Clients were leaving not because of pricing or competitive pressure, but because of service consistency and communication. The loyalty research, such as it was, had been pointing to this for a while. Nobody had acted on it because the findings were inconvenient and the fix required investment in operations rather than marketing. That is a common pattern: research that surfaces an uncomfortable truth gets deprioritised in favour of research that supports the strategy already in motion.

The commercial case for acting on loyalty research is not subtle. Retaining an existing customer costs less than acquiring a new one. Loyal customers tend to spend more over time, refer others, and require less support. The maths are not complicated. What is complicated is building the organisational will to prioritise retention when acquisition is easier to measure and easier to celebrate.

Moz has written thoughtfully about loyalty and the relationship between local businesses and their customers, and while the context is specific, the underlying dynamic applies broadly: genuine loyalty is built through repeated positive experiences, not through programmes or incentives alone. Research that helps you understand where those experiences are falling short is commercially valuable. Research that confirms what you already believe is just expensive reassurance.

Forrester’s research on cross-sell and upsell dynamics is also worth considering alongside loyalty data. Customers who expand their relationship with a brand, buying more products or moving up a tier, are demonstrating a form of loyalty that attitudinal surveys often miss. That behavioural signal is one of the cleaner indicators of genuine commitment to a brand relationship.

If you are building a retention strategy and want to understand where loyalty research fits within the broader framework, the articles in the customer retention section cover the full picture, from programme design to measurement to the commercial levers that actually move retention rates.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions