Value complaints arrive, someone suggests a promotion, and it works — for about five weeks. Then the complaints come back with a new sentence attached: "only worth it with the offer". That sentence is the whole problem, and it is more expensive than the discount that produced it.
Start with the arithmetic, because it is worse than it feels
Take a venue on a 70 per cent gross margin — reasonable for food-led casual dining. A £40 cover contributes £28.
Discount it 20 per cent. The guest now pays £32, food cost is unchanged at £12, and the contribution is £20. You have given away £8 of the £28, which is 29 per cent of your gross contribution to move the headline price by a fifth.
To stand still, you need 40 per cent more covers at the discounted price. Not 20 per cent — 40. And those extra covers arrive on your busiest shifts, because that is when people go out, which means the marginal ones need labour you did not budget and produce the service degradation that generates the next round of reviews.
That is the mechanism nobody diagrams: the discount fills the room, the fuller room slows the service, and slow service is one of the most common causes of value complaints in the first place. You can spend real money making the underlying problem measurably worse.
The deeper cost is that a price becomes a fact
Once a guest has paid £32 for something, £40 is no longer the price. It is the price with a mark-up they now know about. This is not a moral point about loyalty, it is how reference prices work — the last price paid becomes the anchor, and everything above it reads as a premium requiring justification.
Restaurants that discount regularly train a segment of their customers to only appear when discounted. Those guests are not additional. Many of them are your existing full-price guests, rescheduled.
And it shows up in the reviews with characteristic phrasing: "good value on a Tuesday", "wouldn't pay full price", "worth it with the set menu". Every one of those is a review telling future readers what the real price is.
Which value complaint a discount actually answers
There are broadly three, and this is why the instinct misfires:
| The complaint | What it needs | Does a discount help? |
|---|---|---|
| Execution — portions, pace, attention did not match the spend | Operational fix | No. Same experience, less margin to fix it with |
| Communication — surprise charges, unclear menu, mismatched expectations | Wording | No, and it confirms the price was wrong |
| Positioning — priced into a set you are not winning | Strategy | Sometimes, and it is the rarest of the three |
If you have sorted your complaints — and the seven phrasings piece is the sorting method — you will usually find the weight sitting in the first two rows. Those are the two a discount cannot touch.
The four things worth doing instead
Remove the surprises before you touch the price. Bread, water, service charge, sides that are not included. Say it on the menu, say it at the table. This costs nothing, takes an afternoon, and reliably removes a chunk of value complaints, because the complaint was never about the amount.
Add something that costs you 40p and reads as £3. A proper garnish, olives on arrival, decent bread. Guests price the table. The cheapest perceived value in a restaurant is always the small unbilled thing.
Fix the two dishes generating most of your portion complaints. Usually a plating question rather than a quantity one, which is a separate piece. Cost: a crockery order.
If you must do price, do it with a fence. A discount available to everyone resets the price. An offer that is genuinely constrained — a set menu at a specific time, a shoulder-hour proposition, a smaller portion at a lower price — gives price-sensitive guests a way in without telling full-price guests they have been overpaying. The fence is the entire mechanism. A "20% off everything" has no fence.
That last option is also the honest answer to a genuine positioning problem. If your reviews consistently name a category you are losing on price, a smaller cheaper offer is a legitimate strategic response. Just build it as a different product rather than a reduction of the existing one.
The one case for discounting
A new venue with no reviews and no traffic has a distribution problem, not a value problem, and buying trial is reasonable. So is clearing a genuinely quiet shoulder hour where the marginal cover has almost no labour cost attached.
Both of those are fenced, time-limited, and aimed at people who are not currently customers. Neither is what happens when value complaints appear in the reviews and someone suggests doing something about it.
What to do this month
Do not price anything. Sort your last three months of value-flavoured reviews into execution, communication and positioning, and count. If the weight is not in positioning — and it usually is not — the answer is not a number on the menu.
The sorting is the work, and it is the work that stops happening around the second week of a difficult trading month. OMMU keeps the count going in the background and says which of the three is growing, which is a considerably better basis for a pricing decision than the last five reviews you happened to read.