When loyalty looks like loss: measuring the profitability of your loyalty programme

In “How to Keep Hold of Your Customers”, our latest report exploring the subject of shopper loyalty, we discussed three of the biggest questions that retailers need to ask about their loyalty programme. In this follow-up blog, Debora Franchim — dunnhumby’s Director of Loyalty and Personalisation — covers another: “just how do you measure the profitability of a loyalty programme?” 

Thirty years. That’s how long it’s been since Tesco’s Clubcard changed the face of grocery loyalty. Stamps and other basic rewards were around before that, of course, but 1995 was the year in which the possibilities of a data-driven approach to loyalty became truly apparent. Since then, numerous retailers around the world have launched similar initiatives, and loyalty programmes continue to grow both in terms of uptake and sophistication. 

It’s surprising, then, that — even after all this time — the value that loyalty programmes deliver still gets called into question. While the majority of grocery retailers in developed markets now offer some kind of rewards programme, the logic of that decision doesn’t go unquestioned, particularly in tougher economic conditions. Because of that, it’s important for retailers to have a good understanding of just how profitable their loyalty programme really is. 

Problems in the P&L

As with many things in retail, getting to the heart of a loyalty programme’s profitability isn’t as easy as it might first seem. Traditional profit and loss (P&L) accounting is still par for the course within most grocery retail organisations today, something that tends to skew the way in which loyalty programmes are viewed. There are two reasons for that: 

1. Loyalty programmes cost money to run. There’s the immediate outlay, of course, but redeemed points may appear as loss in the P&L, depending on how it’s accounted for It generated misinterpretation of the results. Under traditional P&L models, loyalty programmes typically start life in the red and stay there for the duration.
2. Incremental gains, on the other hand, are not as clear in the P&L as the cost of the program.  Even when a customer buys an item as a direct result of a loyalty programme, there are few ways to prove that for sure, but loyalty-driven purchases tend to get buried in topline sales.

In summary, it’s easy to prove the losses caused by a loyalty programme, and hard to prove the gains. Hardly surprising, then, that loyalty initiatives are sometimes looked on with suspicion by senior execs, not least the CFO. And let’s be clear, challenging investments is entirely healthy. The key is to balance it with a better — and balanced — approach to measurement.

Uncovering loyalty’s hidden value

While the costs of a loyalty programme might occupy a single line on a balance sheet, the true value it delivers is far more nuanced. Even at the most basic level, there are two “kinds” of value that we can attribute to a loyalty programme:

1. The direct, transactional value generated by that programme. Think uplift from redemptions, basket adds, repeat visits, and more. This is value as most people tend to perceive it when it comes to loyalty — the returns they expect it to deliver.
2. The broad and transformational value created by the data that loyalty programmes unlock. Personalisation, insights monetisation, retail media, and pricing strategies are all made possible by loyalty programmes — and their value can be enormous.

There are a couple of important considerations here. Firstly (and fairly obviously), if you’re going to measure the profitability of a loyalty programme then you have to measure both of these streams; you can’t just focus on one or the other. Secondly, it’s important to remember that these “value layers” are linked. Reduce investment into a programme (or even cancel it altogether) and you’re going to lose both the transactional and transformational value that it generates.

A measured approach

Measurement of loyalty programmes is important, then. But how best to do it? At dunnhumby, that’s a question that we’ve spent decades refining the answer to. Today, there are three key ways to measure a programme’s effectiveness.

• Control groups 

Control groups are used to restrict exposure to a programme or activity. Essentially, some customers receive loyalty communications, while others don’t. A comparison between the behaviours of the two groups helps to prove the effectiveness of a loyalty programme. At times it will not measure the entireness of a programme, because some mechanics are available to all customers – such as points distribution – but it is effective to measure the engagement with customers and personalisation that are part of the loyalty programme.

• Comparisons across digital channels

This technique works in much the same way as the above. Some customers use digital touchpoints (like apps and personalised offers), and others don’t. That gives retailers another way to compare behaviours from “exposed” and “unexposed” shoppers. 

• Retrospective control

In specific periods of time, customers who interacted with the programme, especially going to the moments of truth (such as a reward redemption), are separated in a group. Then, a retrospective algorithm identifies customers with the same purchase behaviour in the pre-period. The comparison between the two groups can be considered gains of the loyalty program. The benefit of this approach is not having to hold back a control upfront, but the risk is not finding similar customers if the program penetration is too high.

The halo effect

Regardless of the method, the halo effect must always be considered. The result of a loyalty programme comes from the entire customer behaviour change, not only from a single campaign, activation or product. Say a customer earns points on fresh produce. Over time, they eventually start to buy more from the fresh category overall, not just the products they bought originally. That’s the halo effect, and it’s indicative of long-term, loyalty-driven behavioural change.

Even today — with all of the tools and technologies that we have at our disposal — it isn’t always possible to measure all of those things perfectly. Time and time again, however, we’ve seen that even small steps here can help to build up a better picture of a loyalty programme’s value. And that’s critically important when we consider one of the other big challenges here: continuity.

Trust and turnover

When it comes to loyalty, longevity and trust go hand-in-hand. For a programme to stand the test of time, a retailer’s leadership team needs to trust in its performance. Those teams are subject to frequent change, however — and incoming execs can bring different experiences and expectations to bear on a loyalty programme.

This is another reason that measurement is so important: it ensures (or at least helps to ensure) that the profitability of a loyalty programme doesn’t need to be re-proven every time there’s a change at the top. As I said earlier, challenges are healthy, and these programmes cannot be run on faith. But nor should they be abandoned on impulse — something that a disciplined approach to measurement can prevent from happening.

When times get tough, a loyalty programme can indeed mistakenly start to look like a loss. But cutting it indiscriminately will only ever cost more than it saves. Loyalty isn’t just a scheme or a programme; it’s a long-term strategic asset.