“Why Did She Get That and I Didn’t?”

She saw the text on the other woman's phone

Two customers at the same counter. One of them mentions she came in because she got a message about the new stock landing early. The other one heard about it just now, standing there, for the first time.

"Sorry — why did she get that and I didn't? I shop here all the time."

The associate's options are narrow, and most of them are bad. She can say she doesn't know, which is often true and sounds evasive. She can say it's random, which isn't true. Or she can explain a preview list she's never been told the rules for, to a customer who is about to find out that her loyalty has been assessed and found insufficient.

Everyone else's complaint is about a mistake. This one is about the design.

That's the whole difference, and it's why the usual service moves fail here.

In every other scenario in this series something has gone wrong: a price was mis-ticketed, a delivery failed, a policy was applied badly. The associate's job is to acknowledge, correct where possible, and make it right.

Here nothing has gone wrong. The differential treatment is intentional — a tier, a preview list, a clienteling programme, a personal-shopper relationship, an app segment. It was designed, approved and implemented on purpose, and the customer has just run into it sideways.

Three consequences.

The associate is defending a decision she didn't make and often can't see. Tier criteria usually live in a CRM. Nobody on the floor is told the threshold, so the person being asked is genuinely unable to answer.

"That's just our policy" is much weaker here than usual. In a returns dispute it at least describes a rule applied to everyone. In a retail customer fairness complaint it describes a rule applied to some people, and the customer standing in front of you has just discovered she isn't one of them.

And the complaint is well-founded. She has been treated less well than the person next to her, by design, for reasons nobody has told her. Any response that begins by implying she's mistaken is starting from a false position.

The rule underneath all of this

A tier the customer can't be told about is not a loyalty programme. It's a secret, and secrets produce this conversation.

If the basis can be stated in one sentence, the problem is small. "It goes to members who've opted into messages" or "it's the top tier of the card, which is spend over X a year" — that's a criterion, it's neutral, it's actionable, and most people accept it immediately because there's an obvious route in.

If it can't be stated, the problem is real and it isn't the associate's. Where selection comes from a propensity model, a stylist's personal list, or a segment nobody documented, there is no sentence that helps — and the operator has built something its own staff cannot explain to the people it excludes.

That's the test worth applying to every programme of this kind: could an associate say the basis out loud, at the till, in one line, without embarrassment? Where the answer is no, the exposure isn't a training gap. It's a design decision waiting to be discovered by a customer with a phone.

What to say at the counter

1. Don't deny it. She saw it. Pretending there's no list makes the next ninety seconds much worse.

2. Don't call it random, and don't make it personal. "I think she just knows one of the girls" is the worst available answer and unfortunately a common one.

3. State the basis if you know it, plainly. "It goes out to people on the app who've turned notifications on" — then, crucially, the route: "I can set that up for you now if you'd like."

4. If you don't know the basis, say that, and route it. "I honestly don't know how that list is picked — that's not something I'm told, and I'll find out." Honest, doesn't blame her, doesn't invent a criterion. Then find out, because she'll ask again.

5. Apologise for the surprise, not for the programme. "I'm sorry you found out like this" is true and deliverable. Apologising for the existence of the tier commits the company to something the associate can't give.

6. Give her what you can give. The same stock, held for an hour. The sign-up. A note of her size. Not a substitute for the answer, but it converts a grievance into a transaction.

7. Report the ones that land badly. Nobody upstairs sees these. A note on the day is the only mechanism by which a badly designed tier ever gets revisited.

Design caution. Any exercise built on this should stay strictly on stated, neutral criteria — spend, membership, opt-in — and modules should avoid scenarios where treatment appears to track a customer's characteristics rather than a published rule. That territory carries real legal exposure and belongs with the operator's own counsel, not a training vendor.

Four ways it goes wrong

The denier, who says there's no list. Contradicted by evidence she already has.

The improviser, who invents a criterion. It will be wrong, it will be repeated, and it may be repeated to someone who meets it.

The personaliser, who attributes it to a relationship rather than a rule. Turns a programme into favouritism in one sentence.

The over-apologiser, who agrees it's unfair and implies it'll be sorted — committing the company to abolishing a tier it has just spent a year building.

Why this isn't trained

Loyalty design and floor training are done by different people. Tiers, segments and preview lists are built in commercial and CRM functions. The explanation to a customer who notices is nobody's deliverable.

Associates aren't told the criteria. Frequently deliberately, on the theory that the mechanics are internal — which leaves the person at the till both the most exposed and the least informed.

Clienteling has made it visible. Personal messages, stylist lists and app segments moved differential treatment from the back office to the shop floor, where customers see each other receiving it. The training didn't follow.

And peer role play won't do it. A colleague asked why someone else got early access accepts the first answer. A real customer who has just been ranked in public doesn't, and she asks the second question — what do I have to spend, then? — which is the one that has no comfortable answer.

What tiered-service training can rehearse

A simulation can put the associate in front of someone who has correctly identified that she's being treated less well, and score whether the response stays honest, states a route, and avoids inventing a criterion. Foretell AI supplies the counterparty configuration, transcripts and rubric-based scoring; the tier criteria, what may be disclosed, and the programme design stay with the retailer.

Four to build:

Designing the module

Pass one — the answer. Score whether the differential treatment was acknowledged, whether a criterion was invented, and whether a route in was offered.

Pass two — the unknown basis. Score whether the associate admitted not knowing without blaming the customer or the company.

Pass three — the second question. "So what do I have to do to be on that list?" Score whether the answer was specific or evasive.

Rubric on observable behavior: Was the programme acknowledged? Was a criterion stated, invented, or honestly deferred? Was it attributed to a relationship? Was a route in offered? Was anything promised about changing the programme? Was the incident logged?

Invented criteria are the measure to watch. They're easy to spot in a transcript and they propagate — an associate's guess becomes the store's answer within a week.

The operator case

Undisclosed tiers are a complaint generator with a known cause. The programme works commercially and produces this conversation at a predictable rate, and almost no operator has written the two sentences that resolve it.

The one-sentence test is a cheap design review. Ask whether the basis for every differentiated benefit can be said aloud at the till in one line. The ones that fail are the ones that will surface publicly, and it's better to find them in a meeting.

Clienteling moved this into public view. Programmes designed when communication was private now operate where customers compare notes in a queue. That's a change in the operating environment that most tier designs haven't been revisited for.

And the floor can see which ones land badly. Associates know exactly which segments produce complaints. There is usually no route for that to reach the people who set them, which is a reporting gap rather than a training one.

For retail and customer experience programmes, this is a useful case in the limits of service skill: the conversation can be handled well or badly, but it cannot be made to go away by handling it well, because the cause is a design choice upstream.

Frequently asked questions

How should staff respond when a customer notices someone else got a better offer? Acknowledge it, state the basis if it's known and neutral, offer the route in, and apologise for the surprise rather than for the programme. Never call it random or attribute it to a personal relationship.

What if the associate doesn't know why a customer is in a tier? Say so plainly — that it isn't something they're told and they'll find out. Inventing a criterion is worse, because the guess becomes the store's answer and will eventually be wrong in front of someone who qualifies.

Should loyalty tier criteria be disclosed to customers? If the basis can't be stated in one sentence at the till without embarrassment, that's a signal about the design rather than the disclosure. Stated, neutral criteria with an obvious route in are accepted by most customers; undocumented selection isn't explainable by anyone.

Is differential treatment of customers a problem? Commercially it's standard and it works. The problem arises when it's visible, unexplained and unrouteable — which is what clienteling and app messaging have made far more common.

The short version

She's right, and that's the part nobody prepares for. She was treated less well than the woman beside her, deliberately, by a system that never told her it existed.

Don't deny it, don't invent a reason, don't make it personal. State the basis if there is one, offer the way in, apologise for the surprise, and log it.

And if nobody in the building can say the basis out loud in one sentence, the conversation at the counter was never the thing that needed fixing.

Foretell AI lets retailers build conversational simulations — including tiered-service, loyalty and fairness conversations like the one above — with configurable counterparties, transcripts, recordings, and rubric-based evaluation. If your associates can't state the basis for your preview list in one sentence, we're happy to walk through how other operators have structured it.