The half–year audit most teams are running – and the indicator it misses
Halfway through the year, most marketing, insight and customer teams I work with are running a similar audit. Someone pulls together the headline numbers from the customer base – repeat purchase rate, frequency of engagement, retention by cohort, programme metrics where they exist, NPS if they’ve got it – and forms a view about how things are looking. Often the view is: broadly stable, a few areas to watch. H1 targets are reviewed against forecast. Some are met; in the current climate, often they’re not. The conversation moves to what to focus on in H2.
The audit isn't wrong. It’s measuring behaviour and that’s a useful indicator. But it misses the relationship, the why behind the behaviour.
What the audit is actually measuring
Most customer audits lean on indicators that count what customers have done. Repeat purchases. Engagement with comms. Programme enrolment or redemptions where they exist. Renewal rates. Churn rates.
These indicators are all useful. We use them with a lot of our clients, both setting the classification model up and using them as a first line of enquiry when spend starts changing (typically declining) . They are, however, lagging indicators. They describe behaviour that has already happened, in customer relationships that have already formed and, in some cases, already started to drift.
A customer who keeps buying, keeps renewing, keeps showing up in the dashboard appears loyal in their behaviour but they might also be:
Too busy or lazy to switch
Locked in by a contract, a default, or a points balance they don't want to forfeit
Buying out of habit while their preference has quietly shifted to a competitor
Holding on for one more cycle, waiting to see if you'll fix the thing that's frustrating them
Staying because it's the best price they've found for their budget right now
Loyalty, in the true definition of the word, which is a steadfast commitment, particularly in the face of temptation, is an extremely hard state to achieve with fickle consumers and shoppers. People who are easily distracted by the next best thing, or the brand shouting the loudest or in their face the most.
So, customers who appear loyal are in your retention numbers and showing up as active or engaged, which means the loyalty metrics will keep reporting "fine" until the moment they go.
This is the so–what gap that frustrates so many of the insight and CX directors I speak to. They have ‘the’ number that their organisation has determined is the one to focus on – engagement, retention rate, and so forth. They don't have the why behind it, and they rarely have enough lead time to act on it. The metric is doing exactly what it was designed to do: counting behaviour. Unfortunately that means the metrics are not actually doing what they were intended to do, which is measure and monitor customer relationships; they’re measuring and monitoring behaviour, based on the relationship, probably, 6 months ago.
The cascade you can't see in the dashboard
In every repeat–revenue customer base I've looked at closely, the same cascade typically plays out:
Relationship shifts → behaviour starts to change → engagement declines → spend declines → churn.
The relationship with the organisation is the leading indicator. Behaviour changes come next – smaller basket sizes, paused subscriptions, longer gaps between purchases. Then engagement softens – fewer log–ins, abandoned baskets, less time on platform. Then spend declines. Then churn shows up.
NPS sits in there too, usually between behaviour and engagement. It is a real signal, and a useful one – it's why we build NPS into client trackers. It is, however, rarely the first signal. By the time NPS slips, the relationship has already drifted. The spend drop is normally close behind.
Lagging indicators are still useful – particularly for tracking volumes of customers moving between states (e.g. the new–but–not–converting group growing, or the active–but–disengaging group growing). Those volume shifts tell you where to investigate next. They tell you where to look. They don't tell you what the relationship is, or why it's weakening. Knowing the why is the only thing that gives you the lead time to do something useful about it.
The implication is structural. Most teams are catching the relationship problem three or four steps down the cascade, when the only intervention left is reactive – a discount, a save offer, a winback (and even then, if you don’t know the behaviour is happening, or why, it’s hard to actually have a relevant offer). The earlier you can read the relationship, the more interesting your options become.
This is the case for treating relationships as the leading indicator – not as a soft, philosophical–sounding idea, but as the most practical lens you have on a customer base in a year when budgets are tight and attention is competitive.
What a relationship is actually made of
When we look at relationship health for a client, we're not looking for one thing. A relationship between a customer and a brand is held together by a number of forces, reinforcing each other:
Habit – they buy because they always have
Trust – they've been treated well, especially when it counted
Alignment – this brand fits how I see myself, what I value, who I belong with
Locked–in effort – years of history, data, preferences inside your platform
Value – the deal is good. Held by price, broken by price
Relevance – it meets a need or desire that is still live in my life right now
We don't group customers by which single driver holds them. A relationship may start with one, but a strong relationship is built on a network of multiple drivers. What we do is generate one overall view of relationship health, and then group customers by the strength of that relationship – and then look at the drivers within each group.
So the strongest relationships in a client's base might be held together by trust plus alignment plus habit – customers with a long history, who identify with the brand and don't actively reconsider. The fragile–but–engaged group might be held by habit plus locked–in effort only – they look fine today, and they'll move the moment the friction comes down. The new–and–curious group might be on value plus relevance – good for now, vulnerable to any competitor offering a better deal, or even, quite simply more visible.
This matters because relationships can be deliberately strengthened. A relationship that starts on habit can be reinforced with alignment over time. A relationship that started on value can be moved onto relevance. But a relationship that starts on habit can also be strengthened further with habit, locking it in to their routine. The best customer programmes are proactively designed to do exactly this – they read the relationships they currently hold, and then strategically plan what is most relevant for their brand thereafter.
By understanding the drivers of those unique relationships, brands are better able to give customers what they want, need and what fits into their lives.
By tracking this, they are able to respond proactively, not reactively when the relationship has already started shifting or drifting.
What this looks like in practice
The publications and platforms that have built durable B2C subscription bases over the last few years – the FT, the Times, Disney+ – have, broadly, leant on alignment and relevance. The FT and the Times read like brands you choose because of who you are and how you think; subscribers actively identify with them. Disney+ doesn't rely on inertia; it leans on relevance through release cadence and a deep, identity–aligned content library. None of those bases are immune to churn, but they hold up well in a cost–of–living year because the relationship was never primarily about price.
A smaller example, from a brand I buy from personally is Elavate, a collagen company. My latest delivery had two relationship design moves I've been quietly admiring. The first is a small fridge magnet, marked with the days of the week, designed to be ticked off with each daily dose – a deliberate piece of habit infrastructure. The second is a note on the back of the box: every order fortifies a child's meal for a month, in partnership with the charity Sankus. The first move reinforces habit, critical for a product like collagen to maintain relevance. The second extends alignment onto a product I originally bought for entirely functional reasons. Neither move is dressed up as a loyalty programme. Neither rewards me for already being there. The brand is reading the relationship it currently holds (functional, relevance–based), reinforcing that while deliberately adding relationship drivers that will make it harder to leave when something else arrives.
However, it is worth noting that if they only focused on that – habit and alignment – and didn’t reinforce the main reason I signed up – relevance – they would probably find that the relationship had weakened and was susceptible to competitor activity.
In a year when value–based relationships are the most exposed, this is what good relationship design looks like in operation.
What to do with this in your H2 audit
The honest answer is that most of the data you need is probably already on your team's desks. The lapsed and churned customer conversations you've had this year. The behavioural shifts in the last ninety days that you've noticed but not yet acted on. The patterns hiding in your existing satisfaction and NPS trackers, beneath the headline scores.
You almost certainly have the data in front of you – it's about looking to see what else is hidden in the data, what else is it telling you that you can action right now?
If you're reviewing engagement data, working through H1 results or thinking about a customer strategy reset for the second half of the year, the question worth bringing into the room is not what does our dashboard say? It is: what does our customer base look like by relationship strength, what is each group's relationship built on, and which of those relationships are quietly weakening?
That is the audit most teams are not running. It is also, in our experience, the one that materially changes what gets prioritised in Q3 and Q4.
If that's the conversation you'd like to be having before your second–half plans get fixed, get in touch. We'd be happy to talk through what reading the relationships in your customer base would look like for your organisation.