Case Study 1 — The Value Menu Problem: When the Most Popular Item Is the One Losing You Money
Background
The clearest large-scale demonstration of contribution-margin logic in American foodservice is not in a textbook. It is the twenty-year argument between quick-service brands and their own franchisees over value menus, and it is a matter of extensive public record — earnings calls, franchisee association statements, trade coverage, and litigation.
The pattern is well documented and has repeated across multiple large chains. It runs like this. A brand introduces a national value platform: a fixed low price point, heavily advertised, usually built around the items with the lowest input costs. Traffic rises. Same-store transaction counts improve, which is the number Wall Street watches most closely and the number the marketing department is compensated on.
Then the franchisees — the operators who own individual restaurants and whose income is the profit of one building rather than a royalty stream across thousands — begin to object publicly. Their complaint is consistently the same, and it is a contribution-margin complaint: the value items sell in enormous volume at a contribution margin far below the menu average, they displace higher-margin orders rather than adding to them, and they add labor and throughput cost that the corporate model does not carry.
This has surfaced repeatedly in the public record. Franchisee associations at several major brands have issued open letters objecting to mandated discount pricing. Franchise-agreement disputes over who controls pricing have gone to court. Brands have periodically retreated from value platforms, then reintroduced them under new names when traffic softened. The specific dollar figures inside any given franchisee's P&L are not public, and we will not invent them — but the structure of the argument is entirely public, and it is Chapter 12 at national scale.
The operating issue
Strip away the brand names and the argument reduces to two claims that are both true at once.
The corporate claim: traffic is the business. A guest who comes in for a value item is in the building. They may add a drink, which carries an enormous margin. They may bring three people who order at full price. They may come back. Volume spreads fixed costs, and a restaurant with falling traffic is a restaurant in structural decline no matter what its margins look like this quarter.
The franchisee claim: contribution margin is the business. A transaction is not revenue and revenue is not profit. If the value item's contribution margin is materially below the menu average, and if a meaningful share of value orders are substituting for full-price orders rather than adding to them, then every incremental transaction can make the individual restaurant worse off — while the metric corporate reports gets better.
Notice the shape. This is exactly the plowhorse problem from §12.4 and §12.5, with two additions the matrix does not contain: incrementality (did this order add a guest or replace a better order from a guest who was coming anyway?) and whose P&L we are optimizing (a royalty on gross sales and a profit on one building are not the same objective).
What it shows
First, popularity and profitability are genuinely independent axes, and a business large enough to measure both will eventually discover that its most popular item is not its best one. This is the whole premise of menu engineering, demonstrated at a scale no independent restaurant could produce.
Second, the metric you report is the metric you optimize. Corporate reports transactions and same-store sales. Franchisees live on restaurant-level margin. When the two diverge, the organization does not resolve the disagreement analytically — it resolves it politically, according to who controls the pricing decision. Bellwether's partners will have a smaller version of this argument every single menu meeting, between a chef optimizing for the guest's experience and a partner optimizing for the week's contribution margin. Both are right. Neither is complete.
Third, incrementality is the hardest number in this chapter and nobody has it. §12.5 worked the trout deletion under one set of substitution assumptions and got −$10,273 a year, then changed a single assumption and got +$4,596. The chains have far better data than you will ever have — loyalty programs, order-level histories, market tests — and they still argue about it, because the counterfactual is genuinely unobservable. You cannot see the order a guest would have placed.
Fourth, throughput is a cost the matrix omits. Value platforms lengthen order times and add complexity to the line. In a quick-service model where the constraint is the drive-through, the true cost of a value item includes the cars that did not get served. That is §12.7's labor-and-station-load objection, in the one format where it is large enough to be visible in public financial results.
Outcome
The public record shows no clean resolution, which is itself the finding.
Value platforms have been introduced, expanded, retrenched, and reintroduced across multiple brands over two decades. Franchisee associations have won some pricing concessions and lost others. Several brands have shifted from flat national price points toward bundled offers and app-only discounts — which is a meaningful evolution, because a bundle can be constructed to protect contribution margin in a way a flat price point cannot, and an app-based offer can be targeted at guests who were not coming anyway, which is a direct attack on the incrementality problem.
That shift is the lesson worth taking. The industry's answer to "our most popular item earns the least" was not to delete it. It was to change the structure of the offer so that the popular item pulls something profitable along with it — which is precisely the "attach something to it" move in §12.5's Plowhorse case, scaled to a national marketing budget.
The lesson
High volume at low margin is a strategy, not an accident — and it only works if you have named what the volume is buying you. If the value item buys traffic that converts to attachments, second visits, or fixed-cost absorption, it is doing a job. If it is buying transactions that would have happened anyway at a better margin, it is a transfer from your bank account to your traffic report.
The independent operator's version is smaller and more manageable: know the contribution margin of every item, know which items are subsidizing traffic, and be able to say out loud what that subsidy is purchasing. An operator who cannot answer that question is running a value menu by accident.
Discussion questions
- Bellwether's burger is a Plowhorse: 22.1% of units at a $15.90 contribution margin, $3.16 below the menu average. Is it a value item in the sense described here? What would you need to measure to find out?
- The corporate claim and the franchisee claim are both true. Design a metric that would let a single restaurant tell which one is dominant in its own building this quarter.
- Why is incrementality unobservable even with excellent data? Describe an experiment an independent restaurant could actually run — accepting that it will be imperfect — to estimate it.
- The industry's evolution was from flat price points toward bundles and targeted offers. Explain why a bundle can protect contribution margin better than a discount, using arithmetic.
- In a franchise system, corporate earns a royalty on gross sales and the franchisee earns profit on one restaurant. Show how that structure makes a low-margin, high-volume item rational for one party and irrational for the other. (Chapter 36 returns to this.)
- This case involves brands with thousands of locations and dedicated analytics teams. What, if anything, does a 68-seat independent have that they don't — and how would you use it?