Do not choose a direction before you have a diagnosis
The question is phrased as a choice between three doors, and answering it as a choice is the first mistake. A plateau is an output, and the four common causes lead to genuinely different actions. Getting the diagnosis wrong means entering a new market to solve a retention problem that will follow you there.
You have exhausted the reachable segment, which is genuine saturation. Test it by comparing the addressable population against your penetration of it. If your product serves UK independent pharmacies and there are on the order of a few thousand of them, and you have most of the ones that will ever buy software, then no amount of acquisition spend helps and expansion is the only real answer.
Your acquisition channel has hit a ceiling while the market has not. This is far more common and it looks identical from the revenue line. The signal is that cost per acquisition has risen sharply while win rates on qualified opportunities are unchanged, which says the problem is reach rather than fit. The answer is a new channel, not a new market, and it is a much cheaper answer.
Retention has quietly worsened, so gross additions are flat only because they are being cancelled out. This is the one most often missed, because a headline revenue plateau conceals it completely. Look at new business and churn separately before anything else; if churn has moved, expansion makes the situation worse by adding acquisition cost to a leaking bucket.
Or a competitor is taking the new entrants while your existing base stays loyal. The signal is a normal renewal rate with a falling share of competitive deals, and the answer is positioning rather than geography.
Choose on what transfers
Once the diagnosis genuinely points at expansion, the three options are not symmetrical, and the criterion is which of your existing assets carries over. An asset here means something specific: the product itself, your understanding of the customer, your route to market, your brand's standing, and your operational capability.
| Option | Typically transfers | Typically does not |
|---|
| Upmarket, same buyer type, larger organisations | Product core, domain understanding | Route to market, security and compliance posture, support model |
| Adjacent market, similar product, different buyer | Product core, engineering capability | Domain understanding, brand, channel |
| Deeper into the existing base | Everything | Nothing new is gained if the base is genuinely saturated |
The pattern that this exposes is the one candidates miss: going upmarket keeps your product and discards your go-to-market, while going adjacent keeps your go-to-market only if the buyer is the same person. Upmarket is usually mispriced in planning because the visible work is a feature list — single sign-on, audit logs, a security questionnaire — and the invisible work is a different sales motion with a longer cycle, a procurement process, contractual availability commitments and a support function that answers at night. Teams budget the features and are surprised by the second year.
Going deeper is the option that gets dismissed too quickly, and it deserves a fair hearing precisely because everything transfers. Expanding what you sell to customers who already trust you has the shortest path to revenue of the three, and if the diagnosis was a distribution ceiling rather than saturation, it is frequently the right answer.
Price the option in focus, not in money
The budget for a new market is the least important cost. What decides whether the bet works is attention: whose roadmap absorbs it, which existing customers get less, and how long the organisation has to hold two priorities at once.
Work it visibly. Suppose the team has 200 engineer-weeks a year, and upmarket readiness is estimated at 60 engineer-weeks of compliance and enterprise features. That is 30 per cent of a year, and the core roadmap absorbs the whole of it. If your existing customers are the ones funding the company, you have just given them a year in which the product barely changes, which in a competitive market is how a plateau becomes a decline. That is the argument for sequencing rather than committing: take the smallest slice of the new option that produces a real signal, and protect the core while it runs.
Structure it so you can stop
The bet should be designed backwards from the exit. Ask what the cheapest configuration is that would produce a genuine answer, and what result would make you stop.
For upmarket, the cheap version is winning two accounts of the larger type manually, with the compliance work done as one-off effort and the support handled by whoever is available, before any of it is productised. Two accounts tell you the sales cycle length, the real procurement obstacles, the questions that arrive during security review, and whether the product needs restructuring or only supplementing. That takes a quarter and no platform investment, and the information is not obtainable any other way.
Then name the kill condition in advance and in numbers. Something like: if after two quarters we have not closed two accounts above a stated size, or the average cycle exceeds twice our current one, we stop and put the capacity back into the core. Writing this down before you start is what makes it possible to stop later, because by month nine there will be a salesperson hired for it, a customer mid-implementation, and an executive who announced it. Sunk cost in strategy is organisational rather than financial, and it is much harder to reverse.
The plateau you can enter a new market to avoid discussing
The failure that a candidate cannot see from the outside is that expansion is an attractive answer partly because it postpones a harder conversation. If the real cause is that the product has stopped improving relative to the competition, or that a segment of customers is quietly leaving, then a new market is a way of being busy about growth without addressing either. It is popular internally for exactly that reason: everyone prefers a new story to an unflattering diagnosis.
So the strongest version of this answer includes the case for the fourth option, which is to stay put and fix what is broken. Say what would have to be true for that to be right — churn has moved, win rates against a specific competitor have fallen, or the existing base has unmet needs worth more than the new market — and say how quickly you could tell. A candidate who reaches for expansion without ever considering that the core is the opportunity has answered the question as asked and failed the round.
Diagnose the plateau before choosing a door, then buy the smallest amount of the new market that produces a real answer, with the condition for abandoning it written down while abandoning it is still politically possible.
© 2026 Preptima. Originally published at preptima.com (opens in a new tab).