What GrowSmart Actually Does (In Plain Terms)

Most expansion decisions are made with a familiar shape of evidence: one commissioned feasibility study on the city someone already favoured, a broker's view on two others, and a spreadsheet holding the rest together. The study takes six to twelve weeks and can run to £250,000. By the time it lands, the question has often moved.

GrowSmart is built to change the unit economics of that decision. Here is what it actually is, without the adjectives.

The core object: a verdict, with its working shown

Give it a city and a category, and it returns a single structured analysis containing:

• A demand score (0–100) and an opportunity score (0–100), scored separately because a market can have enormous demand and no room left in it.
• A verdict — Strong Go, Conditional Go, Hold, or No Go.
• Named competitors — the actual operators in that market, not "several incumbents."
• Specific gaps — the capacity band, district and format where supply fails demand.
• Unit economics — capex range, lease benchmark, breakeven occupancy, target occupancy, payback under bull/base/bear.
• Sources, dated, with a Grounded / Mixed / Estimated confidence badge on the analysis as a whole.

That last group is the part that decides whether any of it is usable. A score without its capex assumption is trivia; a score with a breakeven occupancy attached is a decision input.

Four categories, one spine

The framework is constant — demand, competition, location, financials, verdict. What changes per category is the signals inside each dimension:

That calibration is not cosmetic, and the data proves it. Run the same 208 cities through the 5-star hotel lens and the boutique lens and you get near-opposite distributions: 5-star returns 82 No Go and only 5 Strong Go; boutique returns 123 Conditional Go and 40 No Go. Same cities, same engine, different economics — because a market that cannot support a 5-star cost base can often support a boutique one. A single generic "hospitality score" would have averaged that distinction into noise.

There are currently 1,984 cached analyses across 208 distinct cities in seven regions: 155 venue, 416 hotel, and a growing restaurant set.

And a fifth audience. PE funds, REITs and the investors backing any of these operators use the same engine differently — not to pick their own site, but to get a comparable demand-durability read on an asset or platform they're underwriting, in time for the IC memo.

The point of that volume isn't the number. It's that screening stops being a budgeted event. When a study costs six figures, you only commission one for a city you already believe in — which means the process can only ever confirm a shortlist someone assembled by instinct. When the marginal cost of asking about a city approaches zero, you can afford to ask about the twelve cities nobody championed. That is a different decision process, not a faster version of the same one.

What it refuses to do

The engine is instructed not to reach for a comfortable answer, and that holds across every category:

In venue and 5-star hotel alike, more markets are told no outright than earn the top verdict — by roughly four to one and sixteen to one respectively. Mean venue demand score across the index is 59.9, range 22 to 92. This is not a system that returns enthusiasm.

That matters commercially, not just philosophically. A Strong Go is only worth acting on if the same engine was capable of returning No Go on the city next to it. We publish the distribution because a market-intelligence source that has never told you something inconvenient isn't providing analysis — it's providing cover.

Where it fits against a study

The honest framing is that this does not replace the deep diligence you run on the one site you have decided to sign. It replaces the six months of expensive, inconsistent work you do before that — the part that decides which city gets the diligence in the first place, and which is currently done with the least rigour despite carrying the most leverage over the outcome.

What it doesn't do

It won't negotiate your lease, survey your building, or tell you which of two adjacent units to take. It is a market-entry and site-selection instrument, not a substitute for local execution. And it is explicit about its own uncertainty: of the 155 venue analyses, 58 carry a Grounded confidence badge and 96 are Mixed — partially estimated inputs, labelled as such rather than presented with false precision.

Retail is honest about its status too: it runs on the same engine but is still in scoping, so it's stood up on a paid pilot where you see real output before committing, rather than sold as finished.

The takeaway

The pitch reduces to one operational question, and it's the same question whether you operate venues, hotels or restaurants — or underwrite the people who do: how many markets can your current process seriously evaluate in a quarter, and how many should it? If those two numbers are far apart, the binding constraint isn't your team's judgement. It's the cost of forming an opinion at all.

Book a demo and we'll run the engine live on the markets you're weighing right now, on the call, and you keep the output.