You can pass six of the seven growth tests and still miss the number
In November 2025, McKinsey published a litmus test for B2B growth: seven practices its research found in the companies that consistently outgrow their market. It is a good checklist, and it is worth keeping. Here is the summary first, then the part the checklist cannot do for you.
McKinsey's seven tests, in brief
Track against your market. Measure growth versus the market, as granularly as you can, by region, product line, or customer vertical, not just against last year.
Run three to five real growth bets. Each one big enough to add at least 100 basis points of growth, split across your core, your adjacencies, and one or two breakout moves.
Quantify share of wallet per customer. Know the full spend potential of every customer and prospect, not only what they buy from you today.
Earn more than 10% of growth from new customers. Keep a real hunting motion, not just farming the base you already have.
Expand margin about 25 basis points a year. Steady pricing and cost discipline, with increases customers accept because the value keeps rising.
Pay sellers mostly on performance. Commission at least half of total compensation, uncapped, with a wide gap between top and bottom performers.
Make sales tech a force multiplier. An omnichannel experience and a CRM sellers actually want to use, with AI doing real work in the pipeline.
You will read that and nod at most of it. That nod is the problem.
In my experience the plans that fail are rarely built on wrong ideas. They fail because the idea stops at the headline. "Get more clients" is not a plan, it is a wish with a verb. A real plan answers the five questions underneath it: which clients, specifically; why they are not already coming to you; what would actually pull them in; how you would reach them; and against which alternatives they choose today. Skip those and you are not running a strategy, you are funding a slogan.
That is what a checklist quietly hides. Each test is a good headline, and each one assumes you have already answered the five questions beneath it. You have hit most of these marks in some quarter and still watched growth stall, because a mark can be vague enough to check and still too vague to work. The fix is not a longer list. It is the two questions the checklist skips: which test is actually yours, and what order you fix them in.
The seven tests are not seven things
Strip the checklist to its math. Revenue = AOP × NT × ATV × 12. That reads as your average active base across the year, times how often they buy, times what they spend, times twelve. Your base moves by three inputs: customers you add, customers you lose, and what each one is worth.
Those three inputs are the only three ways any business grows revenue. The method calls them the three levers: GET (win new customers), KEEP (hold the ones you have), and GROW (raise what each existing customer is worth).
Now look at the tests again. Test 4, growth from new customers, is the GET lever. Test 3, share of wallet, is the GROW lever. Keeping the base you already have is KEEP. Share of wallet and new-customer growth are not separate disciplines. They are two ends of the same equation. The moment you see the equation, the checklist stops being a list and becomes a choice about which lever carries your year.
Six green lights, one red, and the plan still cancels
Here is what a checklist cannot show you.
The method breaks your marketing into 24 dimensions, the specific parts you can score and fix, from customer experience to pricing to acquisition. Each dimension gets a score from −3 to +3. Some of those dimensions are Fatal Brakes. A Fatal Brake is a weakness severe enough that it cancels the value of everything you are doing right.
Score one at −2 and it does exactly that. You can pass the share-of-wallet test, fund three 100-basis-point bets, and expand margin, and still miss, because your customer Experience (dimension 420) scores −2 and the base leaks faster than you fill it. Six tests green. One red. Plan cancelled.
It does not even take a −2. Run the method on Sage and you see it live. In its strongest year in a decade, subscription penetration at 79% and cloud revenue at record highs, one dimension still held the ceiling down: Experience (420), scored −1, a Fatal Brake for its archetype because the fragmented interface across its legacy products is the opening cloud-native rivals like Xero walk straight through. The record year did not cancel the weakness. The weakness quietly decides what Sage is allowed to become next.
The method has a rule for exactly this. The No-Scale Gate is a simple discipline: you do not fund a growth bet while a Fatal Brake sits unfixed. Not because it is prudent. Because the math cancels the bet before you spend the money. The famous "100 bps bet that never landed" is usually a bet launched over a −2 no one had named. Name it first, and the fix is cheap. Fund around it, and the bet is expensive and the leak is still there.
Which test is even yours
You do not run all seven with equal weight. Your situation names the few that decide the outcome.
The method reads two coordinates. The first is your market's growth curve (M3), the market's clock: is your category just being born, growing fast, mature and flat, or already in decline? The second is your economic value (M4), the depth of what you sell: a bare commodity, a plain product, a service, or a full experience. A steel supplier and a luxury hotel can sit in the same industry and land on opposite ends of that second coordinate. Cross those two answers with the lever you have chosen, and the method returns one of nine archetypes.
An archetype is the strategic pattern your situation fits, the way a doctor matches a set of symptoms to a diagnosis before writing any prescription. Each archetype names the eight dimensions that matter for you and grants permission to stop scoring the other sixteen. If overwhelm is the fear, this is the relief: the method's job is to tell you what to ignore.
Nokia is the cautionary version. At its 2007 peak it held 39% of the global phone market on €51 billion of revenue, and even in 2010 its feature-phone business outsold iPhone and Android combined. That business was declining and commoditising but throwing off cash, a textbook Value Harvester, the archetype for a strong brand in a fading market, whose one job is to grow revenue per customer (dimension 620, ARPU, the average revenue each customer generates) and fund a focused pivot. Test 4 pushes a company like that toward new-customer growth. Nokia's position demanded the opposite. Instead of committing to the harvest, it tried to run three strategies at once, and the method's read puts the cost of that refusal near €100 billion. The checklist would have scored much of Nokia green on the way down. The archetype would have told it to commit.
Run the same equation as a Category Creator, the archetype for a company opening a brand-new market, and every answer flips. Nespresso is the clean case: Nestlé nearly shut it down in 1988, and what saved it was not a better product but naming the right job for the customer and then building a category from zero. Now the story is the strategy, new-customer acquisition is the whole game, and defending the base is next year's problem. Same seven tests. Opposite plan. The difference is not effort. It is which lever your position actually rewards.
Not sure which test is yours?
Answer a few questions about your market and your lever. The quick assessment names your archetype and the handful of dimensions that decide your number, so you know what to fix first and what to ignore.
Take the quick assessment Free to start.The one it will not help you with
Be straight about the boundary. One of the seven, the method will not touch.
Seller commission design (test 6) is a sales-compensation question, not a marketing dimension. There is no honest way to score it on the canvas, so it does not appear. Anyone whose marketing framework claims to fix your comp plan is selling you something. Keep test 6. Just do not expect a marketing system to own it. A method that knows its edges is worth more than one that pretends it has none.
Do one thing, not seven
So take one action from the McKinsey piece, not seven.
Name your lever first: GET, KEEP, or GROW. Then find the single dimension that, left at −2, cancels the rest, and fix that before you fund a single bet. The checklist tells you what good looks like across seven fronts. It cannot tell you which front is yours. That is the whole job of a strategy, and it is the part no litmus test can do for you.
If you want the two coordinates that name your archetype, the finder on the method walks you through it in a few minutes. If you want to see the scoring run end to end, the case library reads more than twenty companies the same way, Nokia and Nespresso among them.
See the scoring run end to end
More than twenty companies, each read the same way: two coordinates, one lever, one archetype, and the eight dimensions that decide the outcome. Same method, opposite conclusions.
