How to look after your top customers: lessons from casinos

Online casinos treat their top customers very differently from the rest, and they learned to do it early: they look at how a new player behaves in the first week after the first deposit, long before total spend tells the story. A small business can copy most of that with a spreadsheet and one phone.

I build retention software for casinos, so I have read a lot of what they do. I will tell you what carries over to a clinic, a shop or a studio, and where I would be careful.

Who counts as a VIP, and why do casinos spot them late?

Most operators define a VIP by totals: a deposit threshold over a month or a quarter. A VIP manager then works from last period's list. A fast-rising player moves through the standard funnel like everyone else, and is noticed after the chance to shape the relationship has passed. Sometimes the player has already left.

The opposite mistake is just as common. A player makes one large deposit and disappears, and gets VIP attention because of a single number. The attention costs real money and does not come back.

Your business has the same two mistakes. The customer who spent a lot last year gets a gift basket. The new customer who has booked three times in a month and asks sharp questions gets a standard receipt.

What do casinos look at in the first week?

They look at behaviour right after the first purchase: the things a good manager would notice by eye, applied to everyone. In casino terms:

• How soon after registering the first deposit came.
• Whether a second deposit followed quickly.
• Bet size, and how it moves over the first sessions.
• How many separate days the player came back.
• The mix of games, and whether the player explores or sticks.

I list these as directions to test and no more. Which of them matters depends on your market and your customers, and your own history answers that better than any rule I could write. The casino version in detail is in how to spot future VIP players in their first week.

For a shop the translation is easy. How soon did the second order follow the first? Did the basket grow? How many different days did they come back? Did they try other product lines? Four columns in a spreadsheet are enough to find out who looks like a future top customer.

Why is a single number a poor guess at someone's value?

Because nobody knows it. In the system I work on, predicted value is shown as a range to day 90, never a single figure. A narrow range on a strong signal deserves a phone call. A wide range deserves a light touch and a second look next week.

You can do the human version. When you think a customer is going to be big, say how sure you are. "Probably a regular, maybe a large one" is a different decision from "clearly a large one". If a vendor shows you one clean value per customer with no spread, ask how they got it.

Who gets a call, and who gets left alone?

Casinos solved this out of necessity, because calls are expensive. The rule is that a call goes to customers whose expected value pays for the minutes. In practice that is the early VIP candidates and the high-value customers who are cooling off. For a mass of low-value customers a call is a poor use of an evening.

Just as important is who gets left alone:

• Customers on a good run. In casinos, a player who is winning right now does not need an incentive to come back tomorrow. In your shop, the customer who just ordered twice does not need a discount.
• Customers whose rhythm says they will return soon. A regular who comes every Friday and has not yet reached this Friday is not lapsed.
• Customers who asked not to be contacted. In a casino, self-excluded players must never receive promotional contact. That is a gate, and no offer overrides it. Treat unsubscribes and "please stop" the same way.

I go into the bonus side of this in who should not get a casino bonus.

Who owns a top customer?

One named person. This is a rule I would add on your side to anything the casinos do: agree who owns a flagged customer, so two people do not contact the same customer and nobody assumes the other one did. It sounds trivial. It is where most VIP programmes quietly break.

How do you keep VIP treatment from going to the wrong people?

Treat early detection as a filter in both directions. The same look at the first week that finds likely top customers shows those whose profile does not support the cost. They do not need to be dropped, only left on the standard path.

One trap is worth naming. A large first purchase says little about value. Casinos have players who deposit only to extract a bonus, who look generous on day one and poor on day thirty. The shop version is the customer who buys with a launch discount, returns the goods and never comes back. Before you invest your time, check that the second purchase happened.

How do you check that VIP attention pays?

The same way as any retention work: compare with a group you left alone. For a few weeks, treat half of your likely top customers as usual and give the other half the call and the personal note. Then compare their next three months. The technique is in how to know a campaign worked. I would be sceptical of any vendor, including me, who quotes a VIP lift before seeing your data.

The engineering version of how casinos score players is in the iGaming retention playbook. You do not need it to start. Everything in this essay can be done by hand.

What can you do today?

Sort your customers by what they spent in the last twelve months and take the top tenth. Write down what share of your revenue they are. That one number tells you how much a lost top customer costs you.

Then look at the ten you know well. When did each last buy, and is any of them overdue by more than their usual gap? Pick the one who is, and call them yourself this week. Skip the offer. Ask how things are going and listen.