Boosting Customer Loyalty: Model, Measurement and Why It Matters

Almost every supermarket chain in the Netherlands now runs a loyalty program. Member numbers are rarely the issue. The real question is whether those members behave differently from customers without a card, and surprisingly often nobody can answer that.

Customer loyalty is the gap between someone who buys from you because you happened to be cheapest that week, and someone who doesn't even check the competitor anymore. For a chain with hundreds of stores, that gap decides how often customers come back and what ends up on the receipt. Which decides your margin.

A conversation I have a lot

A marketing director proudly shows me the member count. Nice growth, big numbers. Then I ask what those members actually do differently from non-members.

There's usually a pause.

Most of the time the program just hands out discounts to people who were already coming anyway. Points get earned and redeemed, but whether that makes someone visit once more a week, nobody knows. The till data sit in one system, the app data in another, and the CRM stands off to the side. Nobody has ever put them side by side. That's where the real work starts: not dreaming up another promotion, but first getting clear on which behaviour you want to change and how you'll see whether it's working.

What customer loyalty actually is

Customer loyalty is the degree to which a customer keeps choosing your brand, even when easy alternatives exist. Researchers split it into behaviour (does someone come back, how often) and attitude (does someone prefer you, do they feel a connection to you).

That split sounds academic. It isn't. A meta-analysis spanning forty years of loyalty programs shows programs mainly shift behaviour, and shift attitude much less. A loyalty point changes what someone buys more easily than it changes what someone thinks of you. Accept that, and you set more realistic goals, and you're less likely to feel let down after year one.

It's also different from satisfaction. A satisfied customer rates one experience positively, but that alone doesn't bring them back.

A lot of Dutch articles on customer loyalty focus on psychology: habits, loss aversion, the IKEA effect. Fine to read. As a marketing or CRM manager, you want something else: how to translate that into something you can actually build and measure.

Why it pays off

Loyal customers come back more often and are harder to lure away with a price promotion two streets over. Retaining them is also cheaper than winning new ones. Research into program benefits shows tangible perks like points, combined with intangible ones like recognition, both feed loyalty toward the retailer.

At the same time, consumers are getting sharper. According to Deloitte, they're looking more critically at what a program actually gives them back.

With chains, the effect shows up fastest. One extra visit per customer per month sounds small, but add it up across hundreds of stores or restaurants and you're talking real money. If you want the basics first, read what a loyalty program is.

A layered model

I like working with a layered model, because it forces you to prioritise.

At the bottom sits the functional layer: convenience, a fair price, products that are actually in stock. Get that wrong and nothing else matters. No points system makes up for an empty shelf.

Above that sits the transactional layer: points and rewards. That drives short-term behaviour, and most programs are fine at it. The top layer is emotional. Recognition and a sense of belonging build attitude over the long run. This is where most brands leave value on the table, often because it's harder to measure.

Research into customer experience and engagement shows engagement is the bridge between a good experience and real loyalty. Without engagement, a good experience stays a one-off. A model that looks specifically at store loyalty adds that trust and consistency over time weigh just as heavily as individual purchases.

Gamification gets added as an extra layer more and more, think a progress bar or levels you climb through. Some research shows this drives repeat purchases. More recent work warns it doesn't always work and can even backfire if the progress is presented poorly. I recognise that. A game bolted on as an afterthought tends to make the program messier, not stronger.

Measuring customer loyalty

You need both sides. Behavioural data tell you what customers do, attitudinal data tell you what they think. The two don't always move together: someone might show up every week for the discount without any real preference for your brand.

I always start with the repeat purchase rate, the share of customers who return within a set period. Churn rate is the mirror image of that. For chains, visit frequency and average basket size are the most direct signals, because you can track them per store and per week. Customer lifetime value shows what a customer is worth over the full relationship. I use NPS as a gauge of preference, not as a goal in itself.

Looking back is useful, looking forward is worth more. Research in omnichannel retail shows you can predict from purchase behaviour which customers are at risk of leaving, before they're actually gone. That only works if your data is in order.

Fashion has an extra wrinkle. A literature review of loyalty programs in fashion shows the effect on retention depends on the balance between personalisation and privacy. Customers want relevant offers. They don't want to feel tracked.

What I'd actually do

Start with personalisation. A study on AI-driven personalisation shows personalised recommendations strengthen trust and loyalty, and McKinsey found that personalisation typically lifts revenue by 10 to 15%.

Segment on behaviour, not just age or location. Two customers with identical profiles can buy in completely different ways.

Reward frequency rather than spend alone. Someone who visits more often is usually worth more to a chain than someone who fills a big cart once in a while, and it's also behaviour a program can genuinely influence. Go easy on gamification too: a progress bar with a clear, reachable goal does more than scattering points everywhere. And don't skip the emotional layer. A birthday reward or early access to a new collection costs little and builds attitude.

For QSR, a tiered program usually works well. For what that looks like, see tiered loyalty for QSR: the McDonald's model. Retail examples are in the best retail loyalty programs, hospitality examples in the best restaurant loyalty programs. An overview by sector is in loyalty program examples across industries.

Keep retention and loyalty separate. Retention is the result, loyalty is often the cause. More on that in what customer retention is.

How NeoDay helps with this

NeoDay is a loyalty platform for B2C brands with many locations, such as supermarkets and petrol stations, but also QSR and fashion chains. It combines AI-driven personalisation with gamification and ready-made frontends. You choose whether it becomes your own app, an SDK inside your existing app, or a web environment. A program can go live within ten weeks, with little work for your IT department.

For IT, what matters most is that the platform is API-centric and cloud-agnostic. It runs alongside your existing systems rather than replacing them. NeoDay won the Best Loyalty Program award in 2025 and is a finalist for the International Loyalty Awards 2026.

For the business case, I'd put scenarios side by side per channel. A customer buying in-store behaves differently from one having something delivered at home, and the margin usually differs too. Keep those apart from the start, and you avoid a program that ends up rewarding the wrong customers. For how the costs break down, see our pricing page.

Looking for quick answers about loyalty programs? Browse the NeoDay loyalty FAQ.