Case: Amazon Buffet — From Confused Multi-Pricing to a National Buffet Leader

A consulting programme with a buffet brand. Its most valuable asset and its most serious problem were the same thing: the category was broad enough that guests could not tell who it was.

Figures here are abstracted to what can be said publicly: guest volumes and qualitative results are published, while detailed dish-level margin and structured data tables are not.

The hard part of buffet is not cost, it is recognition

Buffet has a counter-intuitive property: the more items you offer, the less guests remember.

A buffet restaurant may run hundreds of dishes across stations, and a guest leaving will typically be able to name no more than three. That means the number of items and the brand’s place in someone’s memory are largely unrelated. Repeat visits are decided by a different question: why come to this one specifically, rather than the equally generous one next door.

Amazon Buffet® faced exactly this. It ran a broad multi-category buffet, wide in coverage, but the brand never closed into a single identity. Across stores and service periods the pricing and positioning told guests something different, so the brand did not look like one brand in different places.

Its underlying position was not weak, either: it served more than eight million guests a year, which says the stores’ capacity was never the constraint. What was missing was consistency of identity.

Diagnosis: the pricing was scattered, but the cause was elsewhere

Chen Yufu’s judgement was that multi-pricing was a symptom.

The real issue sat in the competitive frame of reference: without redefining the competitive set, pricing has no single anchor. No single anchor means each store prices against its own trade area — reasonable one decision at a time, and the brand’s identity cut into pieces in aggregate.

That diagnosis set the order of everything that followed. Forcing prices flat as step one would have met store-level resistance, because what each store saw locally genuinely differed. The stores had to first agree on who they compete against; only then does a single price have a basis.

What Yufu did

The programme ran the backbone. Here it is by stage.

Stage	What happened in this programme	What it settled
Product structure design	Established the seafood buffet line inside a multi-category format and reassigned product roles	Gave the brand one point worth remembering
Super menu design	Built salmon as the core product and rebuilt the range around it	Turned positioning into something a guest actually reaches for
Super model building	A Chinese garden style crossed with an international seafood buffet, in space and operations	Made “which restaurant is this” answerable within three seconds
Single-store campaign	One play per store, landing the settled positioning in each	Made head office’s decision actually happen on the floor

Step one was unifying the scattered pricing, with the product structure adjusted at the same time. The two have to move together: unify price alone and stores lose their room to manoeuvre; rebuild structure alone and the identity stays scattered.

Step two was focus. Inside a broad multi-category buffet, the main line was narrowed to a seafood buffet. The cost was explicit — giving up some of the everything-available coverage. What came back was a sense of primary and secondary for the first time: what must exist, what can go, what is only there for company.

Step three was the core product. Around the seafood line, salmon was built into the core product. The test for a core product is not sales volume but whether it can speak for the brand: instantly recognisable, worth a special trip, and something a guest will repeat to someone else.

Step four was expression. The settled positioning was translated into things a guest could carry away: long-running lines about the breadth of the offer, moving toward the idea that for a gathering or a big meal, this is the buffet you go to; plus the brand symbol and a world food map. In space, it landed as a Chinese garden style with an international seafood buffet.

Results

Within the programme, the brand saw:

• pricing unified from a scattered multi-price structure, with the product structure restructured alongside it;
• a seafood buffet main line established inside a multi-category format, with salmon built into the core product;
• the brand symbol and its long-running lines landed, giving the brand a single, repeatable voice for the first time;
• and a position among the leading buffet brands nationally with a Chinese garden and international seafood identity.

These are the brand’s actual results within the programme period, not a promise of operating results to anyone else. Buffet stores are also heavy on floor area and investment, and the rhythm holds: months 1 to 3 lay groundwork, and data usually starts to move in month 4.

What this case really demonstrates

The most valuable move here was being willing to narrow.

The natural temptation in buffet is to add one more thing. Guests like abundance, additions are visible and fast, and cutting items draws complaints almost immediately. So most buffet brands keep adding until the outline is gone.

But abundance is a property of the product, not of the brand. A guest needs one reason to separate this restaurant from the next, and that reason has to fit in a sentence. “We have everything” is not a sentence.

Narrowing to seafood and pushing salmon into the core role trades category breadth for brand clarity. In buffet that trade is especially good value, because guests were never going to remember the full range anyway — they remember the thing that made them choose this one.

Frequently asked questions

Does unifying prices hurt a store’s flexibility? Short term, there is pressure, which is why it has to move with the product structure. Price alone takes away room to manoeuvre; price plus a clearer structure gives the store certainty about what to do.

Why can’t structured data tables be published? Because presenting operating data in structured form is exactly what the anonymisation rules exclude. Comparisons stay at the level of method, or qualitative results with approximate magnitudes.

What does this case share with the Sidaocai programme? In both, the presenting problem was not a bad product but an unsettled competitive frame. That is why both began by redefining the competitive set rather than by changing the menu.