Four tests for a positioning claim

Most positioning claims are compliments a firm pays itself. “Trusted advisor.” “Quality without compromise.” “Results-driven.” Table stakes, wearing a suit. Below are the four tests I run on any positioning claim before I let it near a homepage or a proposal. I use them at Switch, I use them with the CEOs in my Vistage group, and I run them on my own copy, where they hurt most. Two are borrowed and credited, two I’ve assembled from years of workshops. All four are free to steal.

When to run them: before a rebrand, before a website rewrite, or whenever a referral describes your firm back to you and gets it wrong. That last one happens more often than anyone admits.

The opposite test

A positioning claim only means something if a competitor could plausibly claim the opposite. “We deliver excellent work on time” fails instantly, because no firm anywhere says “we deliver mediocre work late.” If nobody would claim the reverse, what you’ve written is the entry fee for your market, and entry fees persuade no one.

The idea builds on David C. Baker, who describes a waterfall of differentiation in his writing on expertise: pour every claim your firm makes down the levels of scrutiny and watch them wash out one by one, because every competitor makes the same ones. Whatever reaches the bottom is your position. For most firms, the bottom is dry.

Running it takes twenty minutes. List every claim from your homepage and your last three proposals. Write the exact opposite next to each one. Then ask, for each pair, whether a sane competitor would say the opposite out loud. If the answer is no, cross it off.

Worked example, invented from scratch: a bookkeeping practice whose site says it is “accurate, responsive, and personal.” All three opposites are absurd, so all three claims get crossed off. The rewrite: “We work on subscription only, call it €2,000 a month, no hourly billing, and we cap the client list at forty so we can answer every query within one business day.” Now check the reverse: “we bill by the hour and take every client we can get.” Plenty of good firms run exactly that model, and proudly. Both sides are defensible, which means the subscription firm has finally said something.

A fail looks like a page of crossed-out lines with nothing left standing. That result means the strategy itself is missing, and sharper adjectives won’t manufacture one.

The stopwatch test

Set a sixty-second timer. Say your positioning out loud, from memory, to someone outside your firm: who you’re for, what you do for them, why you over the alternatives. If the buzzer catches you mid-sentence, or your first sentence needed a caveat, the claim is too foggy to survive contact with a market. Then the harder half. Ask your listener to repeat it back. Their one-sentence version is what your positioning actually is, whatever the deck says.

I once watched the managing partner of an engineering consultancy (a composite, before anyone asks) take four minutes on this, tour the firm’s history since 1987, say “it depends” twice, and get back the summary “so… you do engineering?” That’s a fail. So is any answer that needs slides. The whole test costs one minute plus the swallowing of some pride, and you can rerun it every quarter for nothing.

There’s a reason for the sixty seconds. Most of the time your positioning travels without you: a client describing you at a dinner, a board member forwarding your name with two lines of context. Those people get one sentence each. If you can’t compress the claim yourself, with full knowledge and a running clock, the people repeating it on your behalf have no chance at all.

What would have to be true?

This one comes from A.G. Lafley and Roger Martin’s Playing to Win, and it’s the question I reach for when a leadership team is arguing in circles. Instead of debating whether the positioning is right, list the conditions under which it would be right. What would have to be true about your buyers, about the market, about your own capabilities, about your competitors, for this claim to hold? Then go and check the condition you’d bet on least.

An hour is enough. Four columns, conditions under each, a star next to the shakiest, and the cheapest honest check you can design for each starred item. Five phone calls beat a survey every time.

Worked example, again invented: a software shop wants to claim “the fastest dependable delivery partner for mid-market logistics companies.” For that to hold, operations directors at logistics firms must rank delivery speed above price when they pick a partner; the shop must be able to staff a new project within two weeks in any month of the year; and the big integrators must be unwilling to match that speed at mid-market budgets. The first condition is the shaky one. Ten calls to operations directors would settle it inside a week, and either result is useful.

A fail is quieter here than in the other tests. It’s the condition everyone in the room privately knows is false, left on the list, with the claim shipping anyway. The value of the question is that it turns an opinion fight into a research task. Nobody has to win the meeting.

The sameness scan

Collect the About pages of your five nearest competitors. Paste them into a single document with your own, strip out names, logos, photographs, and city references, and hand the anonymised file to someone who knows your firm well: a new hire, say, or a friendly client. One instruction. Find us.

Thirty minutes to collect, ten to run. What it exposes is category wallpaper, the copy every firm in a market ends up sharing because everyone wrote theirs while looking at everyone else’s.

Worked example, invented: six commercial law firms. Five of the six paragraphs contain “client-focused,” “decades of combined experience,” and some arrangement of “practical, commercial advice.” Interchangeable. The sixth names the two kinds of work the firm refuses to take and prints fixed fees for its most common matters. It even tells you who answers the phone. That paragraph survives with the logo stripped away, and it’s the only one that does.

A fail looks like your own operations lead reading the file slowly and then, with complete confidence, picking a competitor’s paragraph as yours. I’ve run this in workshops and watched it happen. The room goes quiet in a very useful way.

When you fail it, resist the urge to write more. The fix is subtraction: cut every sentence that could sit on a rival’s page without anyone noticing, then look at what survives. Usually it’s two sentences. Usually they’re the true ones. Rebuild from those.

Run them in this order: the sameness scan first, because it’s the cheapest and it stings fastest; the opposite test second; the stopwatch third; and “what would have to be true?” last, because it’s the only one that generates homework. An afternoon covers the first three. Budget a week for the calls the fourth one demands.

Credit where it belongs: the opposite test grew out of David C. Baker’s waterfall of differentiation (his book The Business of Expertise is worth your time), and “what would have to be true?” belongs to Lafley and Martin. The stopwatch test and the sameness scan came out of my own workshops, though I doubt I’m the first person to time an executive or redact a logo. If you know who was, tell me and I’ll credit them here.

— Rik