Case: Sidaocai, a Fujian Restaurant Brand — From Two Loss-Making Regions to Profit

A year-long programme with a Fujian cuisine brand. The problem was not that the food was wrong — it was that two regions were losing money at once, and head office was applying one set of actions to two very different situations.

Figures here are abstracted to what can be said publicly: store counts and qualitative results are published, while full P&L statements, dish-level margins and internal names are not.

Why Fujian cuisine, and not another category

Sidaocai was founded in Fuzhou in 2009 — the home of Fujian cuisine, one of China’s eight great culinary traditions. The style rests on fresh ingredients and original flavour: Fujian ingredients cooked as Fujian food.

That origin sets two things: the radius of the supply chain and the distinctiveness of the taste. The brand later expanded beyond the province into Jiangsu, Anhui and Shaanxi, reaching more than 50 stores across six provinces, and collected a series of public recognitions including municipal Fujian-cuisine restaurant listings, culinary guide listings and repeated appearances on Dianping’s must-eat lists.

The growth story reads smoothly, but the problem was hiding inside it. Stores were opened one at a time, and each new city brought a different competitive set, customer base and pricing environment. By the time the regions took shape, head office was holding a playbook that local experience had already cut into pieces.

Diagnosis: why the competitive set came first

Chen Yufu’s judgement was that changing the menu at this stage would be the most dangerous move available.

The reasoning: a menu answers “how do we sell”, but the unresolved question was “who are we compared against”. Until the competitive set is redefined, there is no anchor for pricing; without a pricing anchor there is no basis for deciding which dishes stay; and so a menu is rebuilt, budget is spent, and the problem simply moves to the next quarter intact.

So the first step was redefining the competitive set: in each market, who does the customer actually weigh this brand against? Once that was settled, two things became clear at the same time — which regions had a positioning problem and which had a product problem.

Those two problems need two different sets of actions. Run them together and neither gets done.

What Yufu did: the four-stage backbone

The whole programme ran on one backbone. Here it is in delivery order.

Stage	What happened in this programme	What it settled
Product structure design	Reassigned dishes to signature, base, pairing and traffic roles	Decided what each dish is for, not just whether it tastes good
Super menu design	Translated the settled positioning into the page a guest actually reads	Made the structure work at the point of ordering
Super model building	Fixed the working actions into a repeatable single-store model	Meant the next store did not start from scratch
Single-store campaign	One play per store; loss-making stores handled as different single-store models	Decided which store to turn and which to stop

Product structure design produced a list of why each dish exists, not a new menu. Signature products carry recognition, base products carry repeat visits, pairing products lift the average check, traffic products bring new customers. The four roles tolerate completely different costs, station loads and margins — manage them as one group and they drag each other down.

Super menu design followed: once the structure is set, the guest has to be able to read it.

Results

Within the programme, the brand saw three things:

• Both regions moved from loss-making to profitable, with growth above 30 per cent;
• with the competitive set redefined and the positioning upgraded, a brand symbol landed and gave the brand a mark recognisable on storefronts and materials;
• the product structure was rebuilt and blended margin across stores improved within a range.

Worth stating plainly: these are that brand’s actual results within the programme period. They show the chain worked for this brand, not a promise of operating results to anyone else. Yufu’s consistent position is also that months 1 to 3 lay the groundwork and the data usually starts to move in month 4 — we do not promise a faster pace.

What this case really demonstrates

The step that mattered most here was also the easiest to miss: when both regions were losing money, the team did not rush to fix the product.

With two regions bleeding, the pressure sits entirely on head office, and the natural reflex is to reach for something that shows results fast — dishes, price, promotion, pick one and ship it. Yufu’s team chose positioning instead. That is not slowness; it is avoiding work that would have to be undone.

When one brand faces two regions in different situations, the correct order is to settle the “who are we compared against” layer first, then let each region configure itself inside that positioning. Get the order right and the same resources produce two results. Get it wrong and the same resources cancel each other out.

Frequently asked questions

Why can the brand name be published here? Brand names can be disclosed, which is the agreed position with the client. What stays out is detailed margin, structured data tables, full P&L statements and the names of client-side staff — client people are referred to by role only.

Why is there no operating data table in the article? Because presenting operating data in structured form is exactly what the anonymisation rules exclude. Comparisons are made at the level of method, or as qualitative results plus approximate magnitudes.

How long is the programme? Sidaocai ran the annual strategy consulting service. Yufu’s service lines move in phases — the Strategic Positioning Program for the first six months, then Enterprise Strategy Consulting across a cumulative year: two phases of one backbone, not two separate products.