VIP customers: how to identify them and what to actually do for them

In most shops a small minority of customers produce a large share of revenue, and the gap between that group and everybody else is wider than any demographic split you could construct. Treating them identically to a first-time visitor is the default behaviour of nearly every marketing setup, and it costs money in both directions at once: you under-serve the people who pay you most, and you over-discount the people who would have bought anyway.

A VIP programme is not a loyalty card and it does not need points, tiers with names, or a plugin. It is a segment with different rules attached, and the hard parts are defining it honestly and deciding what the reward actually is.

Defining the group without using lifetime spend

Lifetime spend is the obvious definition and it has two serious problems. It never decreases, so a customer who stopped buying two years ago remains a VIP indefinitely. And it rewards a single large purchase identically to a durable relationship, which means your top segment fills up with people who bought one expensive thing once and have no particular attachment to you.

1. Use a rolling window. Revenue over the last twelve months, recomputed monthly, so the group reflects who is valuable now rather than who was valuable once.
2. Require recency independently. A high-value customer who has not ordered in a year belongs in win-back, not in VIP, however much they spent in month two of the window.
3. Use a relative threshold — the top five or ten per cent — rather than an absolute euro figure. An absolute threshold set this year is wrong next year, and nobody ever remembers to revisit it.
4. Compute net of returns and refunds. This is the one that surprises shops. A customer who orders constantly and sends half of it back is expensive, not valuable, and several shops discover on first running this that a visible fraction of their apparent top segment is exactly that.
5. Consider frequency alongside value where your margins vary by product. Two customers who spent the same amount, one across six orders and one in a single purchase, are different people with different futures.

None of this requires a model. It is four columns and a percentile, computed from orders you already have, and it will be more accurate than any scoring plugin because it uses your real refund data.

Why the reward should not be a discount

Discounting your best customers is paying the people most likely to buy at full price not to.

These are, by definition, the least price-sensitive people on your list. They have already bought repeatedly at your normal prices, which is the strongest possible evidence that your prices are acceptable to them. A standing percentage off reduces your margin on the revenue you were most confident of, and it teaches a group that was not discount-driven to start waiting for the code — which is a behaviour change you cannot easily undo.

What works instead costs less and is considerably harder for a competitor to copy, because it is about service rather than price:

• Early access. New arrivals and sale periods, a day or two ahead of everyone else. Costs nothing, is genuinely valued, and has the useful side effect of clearing stock before the general announcement.
• Free or upgraded shipping as a standing benefit. Predictable, easy to explain, and it removes friction on precisely the repeat orders you want more of.
• A real person to contact when something goes wrong, answered faster. For a small shop this is simply a reply-to address that reaches somebody.
• Extended returns or a no-questions exchange window. Low cost for a group that returns less than average, and it removes hesitation on higher-value orders.
• Genuine recognition. A note, a small extra in the parcel, or simply being told they are in the group. The cheapest item on this list and the one most often skipped.

The moment somebody enters the group

The entry message is the one email in this programme that matters most, and in most shops it does not exist at all — benefits are applied silently and the customer never learns about them.

A benefit nobody knows about changes nobody's behaviour. It is a cost with no return.

1. Tell them, explicitly. This is the whole point.
2. Say why, concretely. "You have ordered six times this year" is a reason. "You are a valued customer" is a greeting card. The specific version is both more credible and more flattering, because it demonstrates you noticed.
3. List what changes, plainly, with no expiry date attached. A benefit with a countdown on it is a promotion wearing a different hat.
4. Ask for nothing in this message. No offer, no upsell, no survey, no referral request. It is the one email in your entire programme that gives without asking, and that asymmetry is exactly what makes it work. Put the ask in the next one.
5. Send it from a person, with a reply-to that reaches somebody. A meaningful share of recipients will reply, and those replies are the most useful customer research you will get all year.

The harder problem: people who fall out

In a rolling model, somebody whose trailing twelve-month revenue drops below the threshold has left the group. Handling that badly is worse than never having run a programme at all, because you have now drawn attention to something being taken away.

• Never send a demotion email. There is no wording of "you are no longer a VIP" that reads well, and the attempt to soften it makes it worse.
• Let benefits lapse quietly, or give a grace period of a few months. Nobody audits their own shipping benefits.
• Treat the drop as a signal rather than a status change. A customer sliding out of your top decile is the single most valuable at-risk person on your list — proven high spend, currently declining — and they should enter a retention sequence well before the threshold is actually crossed.
• Watch the leading indicator, not the crossing. A VIP whose order interval has stretched by fifty per cent is leaving; the revenue figure will confirm it four months later.
• Report the flow between groups monthly. How many entered, how many left, and the net. That number moves before revenue does and it is the earliest warning you will get that something has changed in product, delivery or competition.

The operational side nobody mentions

A VIP programme touches more than email, and the parts outside marketing are where it either becomes real or quietly stops working.

1. Support has to know. If your helpdesk cannot see that the person writing in is in the top decile, the "faster answer" benefit does not exist. This is usually a one-field integration and it is the highest-return part of the whole programme.
2. Shipping rules have to apply automatically. A free-shipping benefit the customer has to remember to claim with a code is not a benefit, it is a quiz.
3. Stock decisions should reflect it. If your best customers are concentrated in particular products, those are the products not to run out of.
4. Keep the group out of acquisition advertising. Paying to retarget your most loyal customers is the most avoidable waste in the entire marketing budget, and it requires only a suppression audience refreshed from real orders.
5. Decide what happens to a VIP who returns a lot. The net-of-returns definition handles it automatically, which is the argument for computing it that way in the first place.

Measuring it honestly, which takes a year

VIP programmes flatter themselves more than any other flow, because the group was selected for buying a great deal and is then credited with buying a great deal. The segment's revenue will look spectacular whatever you do, including nothing.

The only honest test is a holdout, and it is a slow one. Withhold the benefits from a random slice of newly qualifying customers and compare their next twelve months against the rest of the cohort. It needs a year to read, which is why almost nobody does it — and why almost nobody can tell you whether their loyalty programme does anything.

Three things worth reporting in the meantime:

• Retention rate of the group year on year, which is the outcome you actually care about.
• Flow in and out monthly, as the early warning described above.
• Share of total revenue from the group, watched for concentration. If it climbs past a point you are comfortable with, that is a business risk rather than a marketing success, and the right response is acquisition rather than more VIP benefits.

That last point is the one worth sitting with. A programme that works perfectly makes your revenue more concentrated in fewer customers, and concentration is fragility. The VIP flow and the new-customer flow are in tension, and a shop should know which way it is currently leaning.