The slide had nineteen flags on it. I counted them while the founder talked, because the flags were doing something the revenue numbers were not: they were creating the impression of a company that had cracked international.
Then we opened the pipeline. Four markets produced 92% of closed revenue. India, the UAE, the US and Singapore. The other fifteen flags were one pilot each, mostly sourced through a conference, mostly stalled at the second meeting, all of them consuming somebody’s calendar. Worse, the four real markets were being served by a single positioning statement written eighteen months earlier for a US audience, then translated, then softened, then softened again until it said nothing offensive to anyone and nothing useful to anybody.
That is the actual condition of most companies that describe themselves as multi-market. Not twenty countries. Four markets, served badly, by a message engineered to survive all four rather than win any of them.
The compromise message is the tax you pay for not choosing
Here is what happens mechanically. You start with a sharp claim, the one that won your first fifteen customers. Then a deal in the Gulf stalls because the buyer there does not care about the metric you lead with. So you broaden the claim. Then an Indian prospect pushes back on price and you learn they are buying against a completely different alternative, so you add a line about efficiency. Then a US enterprise buyer asks about security posture and you promote compliance into the headline. Six iterations later your homepage says you are an intelligent platform that helps organisations optimise operations and drive growth.
Nobody wrote that sentence. It accreted. And it is expensive in a way that never shows up in a P&L, because the cost is measured in meetings that go one round longer than they should, in sales cycles that stretch from ninety days to two hundred, in a product roadmap that receives contradictory signals from four directions and resolves them by building shallow versions of everything.
I watched this from the inside at HSBC in Hong Kong, working on machine learning in lending and personalisation. A bank operating across dozens of markets does not have one lending product. It has a portfolio of businesses that share a brand and a balance sheet, and the discipline that keeps it coherent is not messaging discipline. It is knowing precisely which risk committee, which regulator and which budget line owns each product in each market. The story follows the money. It is never the other way round.
Most founders try to do the reverse. They write the story first and hope the money arranges itself behind it.
Budget authority is the border. Language and flags are not.
Here is the reframe that changes the whole conversation. Stop segmenting by country. Segment by which budget the money comes out of.
Two markets that feel completely different culturally can be the same market for you, if in both cases the buyer is a Chief Risk Officer spending a compliance budget approved annually, defending the purchase to a regulator. The sales motion is the same, the proof points are the same, the objection handling is the same, the price anchor is the same. Language differs. Business model does not.
And two markets that look adjacent on a map can be two entirely different companies. Selling into a Gulf entity where the buyer is a transformation office spending a capital allocation tied to a national programme, with a mandate to show visible modernisation, is a fundamentally different business from selling into an Indian mid-market operator where the buyer is a COO spending from operating budget, measured on cost per transaction, negotiating hard because their own customers negotiate hard with them. Same continent-ish. Same sector. Same product, technically. Not the same company.
The transformation office buys a story about capability and ambition, over a longer cycle, at a price that reflects strategic significance, with success measured in visibility and reference value. The COO buys a story about unit economics, over a shorter cycle, at a price that has to clear an internal payback threshold, with success measured in a number that appears on a monthly report. If you try to serve both with one deck, the transformation office finds you small and the COO finds you expensive.
This is the test I now apply before anything else. Do these two markets buy from the same budget line? If the answer is no, they are two companies. Two price points. Two proof libraries. Two sets of reference customers. And eventually, once volume justifies it, two owners with separate targets.
Everything people usually agonise about is downstream of that one call. Pricing tiers are downstream. Payment rails are downstream. Local entity structure, invoicing currency, data residency, partner versus direct, all downstream. Founders spend months on the downstream questions because they are concrete and answerable, while the upstream question stays open because it requires giving something up.
Localisation copy is not a fix. It is a coat of paint on a structural problem.
The reflex, when the message stops working in a new market, is to localise. Translate the site. Hire a regional marketing contractor. Change the case studies on the country page. Swap the stock photography.
I have never seen this move a stalled market, and I have watched it attempted several times. The reason is simple: the message was not failing because of language. It was failing because it was answering a question the buyer had not asked, in a currency of proof the buyer does not accept.
When we were building consumer hardware in China for the venture with Jean-Michel Jarre, we grew one product into a line of eight. Those eight were not translations of each other. They existed because different buyers had different jobs, different price sensitivities and different retail contexts, and a single product stretched to cover all of them would have been mediocre everywhere. Hardware forces this honesty on you, because you cannot ship a product with a toggle that makes it simultaneously premium and value. Software lets you pretend, and the pretending is what kills positioning.
Later, running digital transformation for a private healthcare group expanding across mainland China, the same lesson arrived in a different costume. Tier one city hospitals and inland facilities were nominally the same customer inside the same organisation. In practice they had different capital approval routes, different staffing realities and different definitions of a successful rollout. Treating them as one deployment produced a system that satisfied neither. Splitting the approach cost more upfront and worked.
Choose the market that owns the roadmap. Let the others be opportunistic.
This is the part founders resist, so I will state it plainly. One market gets to set the product roadmap. The others are allowed to generate revenue, and they are not allowed to generate requirements.
Pick the market where the buyer profile is clearest, the sales cycle is most repeatable, and the willingness to pay is highest relative to your cost to serve. That market owns the backlog. Its feature requests go into the plan. Its language sets the primary positioning. Its reference customers appear first.
The other markets stay open. You take the deals that come, you serve them properly, and you price them from the standard tier structure rather than negotiating a bespoke arrangement each time. What you do not do is build for them. A feature request from an opportunistic market goes into a log, and it only leaves that log if three unrelated buyers in that market ask for the same thing inside two quarters. That threshold is the whole discipline. Without it, a single enthusiastic prospect in a market you have no strategy for can consume a quarter of engineering capacity.
When I co-founded a location-based platform in China and it grew past 300,000 users, almost every meaningful improvement came from a narrow understanding of one user context, deeply. The temptation to generalise early is enormous and it is almost always wrong. Depth in one context produces something people love. Breadth across four produces something people tolerate.
There is a version of this that is uncomfortable to say out loud in a board meeting, which is that some of your revenue is strategic and some of it is just money. Both are fine. Confusing them is not. Money that is just money should never be allowed to vote on the roadmap.
What to do this quarter
Take your closed revenue for the last four quarters and re-cut it, not by country, but by which internal budget paid for it. You will typically find two or three clusters, not nineteen. Name them by buyer and budget, something like “risk budget, regulated, annual approval” and “operations budget, cost-justified, quarterly approval.” If two of your markets fall into the same cluster, congratulations, they are one business and one message serves them.
For each distinct cluster, write the positioning statement that would win with no compromise at all, as though that cluster were your only customer. Then look at how far apart those statements are. If they are close, you have a real single-market story and your problem is execution. If they are irreconcilable, you have your answer, and the honest next step is to name the primary, set the threshold rule for everything else, and stop pretending the compromise sentence on your homepage is doing any work.
Twenty flags is not a growth strategy. It is an unmade decision, formatted attractively.
I write from twenty years of building businesses between Europe and Asia. If your company is facing this, start a conversation.