Costs go up, so you think about raising prices. Then the doubts start. Will the regulars leave? Will the shop next door look cheaper? So the price stays where it is, and the extra cost comes out of your profit.
That feels safe. It isn't free. Holding a price is a decision too, and it has a cost you can calculate.
Why gut feel gets it wrong
Gut feel treats a price rise as a bet you could lose. It counts the customers who might leave. It forgets that every customer who stays now pays more.
Those two effects are never put side by side. Doing that takes one number: your margin on the product.
What big companies do: treat price as the biggest profit lever
In The power of pricing (McKinsey Quarterly, 2003), Michael Marn, Eric Roegner and Craig Zawada took the average income statement of an S&P 1500 company. A price rise of one percent, if volumes stayed stable, would lift operating profit by 8%.
| Exhibit 1, average S&P 1500 company (index) | Before | After a one percent price rise |
|---|---|---|
| Revenue | 100.0 | 101 |
| Variable costs | 68.3 | 68.3 |
| Fixed costs | 19.2 | 19.2 |
| Operating profit | 12.5 | 13.5 |
On the same statement, our arithmetic from Exhibit 1 shows that a one percent fall in variable costs lifts profit by 5.5%, and a one percent rise in volume lifts it by 2.5%. Price is the strongest lever.
The article also shows the other side. Volumes would have to rise by 18.7% just to offset the profit lost by a five percent price cut.
But the authors are clear that this is not about raising prices across the board. They write that the most effective path is often "to get prices right for one customer, one transaction at a time, and to capture more of the price that you already, in theory, charge." Their tool is the pocket price waterfall: every discount, rebate, free delivery and payment-term giveaway between the list price and the cash you actually keep.
One example in the article, a disguised lighting supplier, shows what that looks like. On-invoice discounts took its prices 32.8% below list, and off-invoice items took another 16.3 points, so the pocket price was 50.9% of list. After it brought discounts into line and enforced its rules, its average pocket price rose 3.6% and operating profit rose 51% in the first year. McKinsey disguised the client, so treat this as an illustration, not proof.
What we changed for small businesses
McKinsey's 8% is an average for big US listed companies. It tells you the stakes, not your answer. Your answer needs your own margin. So this version turns the arithmetic around and asks one question: how many customers can I lose before the rise stops paying?
Margin = (price − direct cost) ÷ price
Sales you can lose = rise ÷ (margin + rise)
Both are percentages of today's price and today's sales. The second line is the same arithmetic as the 18.7% above, run in the other direction.
| McKinsey's version | SME version | Where you find it |
|---|---|---|
| The average income statement of a listed company | Your own margin on each product | Your costing, POS or invoices |
| Pocket price waterfall | A leak list: free add-ons, standing discounts, platform commissions, free delivery, rounding down | Last month's invoices and orders |
| Volume assumed stable | The volume loss you can afford (the break-even) | The formula above |
Try it with McKinsey's own statement. Variable costs are 68.3 of every 100 of revenue, so the margin is 31.7%. A one percent rise survives a volume loss of up to 1 ÷ (31.7 + 1) = 3.1%. A five percent rise survives a loss of up to 5 ÷ (31.7 + 5) = 13.6%.
Honest limits: the formula assumes your fixed costs don't change. It tells you how many sales you can afford to lose, not how many you will lose. That judgement is still yours. And a customer who leaves also stops buying the drink, the dessert and the add-on, so lean towards caution.
What to do in each case
Put the line next to the loss you honestly expect.
| Your situation | What to do |
|---|---|
| The loss you expect is well inside the line | Raise it. The rise pays even after some customers leave. |
| The loss you expect is close to the line | Test it. Raise one product or one group of customers first and watch for a month. |
| The loss you expect is beyond the line | Hold it. Fix leaks and costs first, or add something that supports the price. |
| You don't know what customers will do | Find out small. One product or one outlet, then count orders. |
One framework, three businesses
The same arithmetic works whatever you sell. These are illustrative examples, not real businesses:
- A café. Example with made-up numbers: a set meal sells at RM10.00 and costs RM6.00 in ingredients and packaging, so the margin is 40%. A 5% rise (RM0.50) pays as long as sales fall by less than 5 ÷ (40 + 5) = 11.1%. Before the rise, look at the free teh ais that comes with every set.
- A B2B supplier. Example with made-up numbers: a margin of 20% and a 3% rise give a line of 3 ÷ (20 + 3) = 13.0%. A thin margin gives a wider line. The first move is not the list price. It is the standing discount that long-time buyers get and newer buyers don't, the same pattern as in McKinsey's lighting supplier.
- An online shop. Example with made-up numbers: a margin of 50% and a 5% rise give a line of 5 ÷ (50 + 5) = 9.1%. A fat margin gives a narrower line. Check free-shipping thresholds and voucher codes before you touch the price tag.
A real case: Cracker Barrel, fiscal 2023
Cracker Barrel is a US casual-dining chain, far bigger than any SME. It is listed, so its annual report (Form 10-K) says what it did to menu prices and what happened to traffic. This is its fiscal 2023, the year ended 28 July 2023, against the year before.
| Measure | Result |
|---|---|
| Average menu price | +8.6% |
| Average check (menu price plus mix and items per guest) | +9.8% |
| Guest traffic | −3.5% |
| Comparable-store restaurant sales | +6.3% |
| Comparable-store retail (gift shop) sales | −0.4% |
| Food cost, as a share of restaurant revenue | 28.1%, up from 27.5% |
| Break-even traffic loss on food cost alone (our arithmetic) | 10.6% |
The break-even is 8.6 ÷ (72.5 + 8.6) = 10.6%. Food cost was 27.5% of restaurant revenue in the year before, which leaves 72.5% to cover everything else. We treat food as the only variable cost.
What it shows:
- The loss sat well inside the line. Menu prices rose 8.6%. Traffic fell 3.5%, against a break-even of 10.6%. Across all its restaurants, restaurant sales less food cost grew 6.03% (our arithmetic from the 10-K).
- The rise did not fully pay for costs. Food cost still climbed from 25.5% to 27.5% to 28.1% of restaurant revenue over two years of price rises.
- Lost visits cost add-on spend. Retail sales fell 0.4%, which the company puts mainly down to fewer guests. Your break-even should count the add-ons too.
Caveats you should keep in mind:
- Traffic has many drivers, and the 10-K does not blame the 3.5% fall on the price rise alone. Read it as what happened in the year prices rose 8.6%, not as a measured reaction to price.
- This is an illustration of the arithmetic. It is a US chain with 659 comparable stores, not an SME and not a Malaysian benchmark.
- The 10.6% is ours, not the company's. It uses food cost as the only variable cost and the year-earlier ratio. Counting hourly labour as variable cost would give a thinner margin and a wider line, so 10.6% is the cautious end.
- The 6.03% covers all restaurants, while traffic is for the comparable stores. Putting them side by side is approximate.
- In the year before, menu prices rose 5.9% and traffic rose 8.0%. That was a rebound year, so we don't compare the two years.
Try it on your own prices
You need no tool, only last month's numbers. Do this for your top 3 to 5 products.
- Work out the margin. (Price − what you pay for each unit sold) ÷ price. Count ingredients, packaging, card and platform fees.
- Write the leak list. Free add-ons, standing discounts, vouchers, platform commissions, free delivery, rounding down. Add up what each cost last month.
- Plug the leaks that no customer will see. McKinsey's first lever is to capture the price you already charge.
- Work out the line for the rise you have in mind: rise ÷ (margin + rise).
- Write down the loss you expect. Be honest. Look at what happened to orders the last time you changed a price.
- Decide with the table above. Start with one product, then count orders and basket size for four weeks.
Three rules before you act
- Plug leaks before you touch the menu price. The cheapest rise is the one customers never see.
- Use your own margin. Don't borrow McKinsey's 31.7%. It comes from someone else's accounts.
- Count what happens. Track orders and basket size for a month, and be ready to roll one item back.
The break-even doesn't tell you what customers will do. It tells you how wrong you can afford to be.
Sources
- Michael V. Marn, Eric V. Roegner and Craig C. Zawada, The power of pricing, The McKinsey Quarterly, 2003 Number 1, pp. 27–36: Exhibit 1 (p. 29), lighting supplier (pp. 31–33). Copy hosted on a University of Colorado course page.
- Cracker Barrel Old Country Store, Inc., Form 10-K for fiscal 2023 (year ended 28 July 2023): Item 7 MD&A, pp. 34–35.