Deep market research, weighed.
Also called: Market sizing · Competitor mapping · Demand validation · Quantitative market analysis
The deeper, numbers-led market work: sizing the real addressable market, mapping competitors in detail, and proving demand before you commit money.
Size the addressable market, map every competitor, and test price and demand with evidence, not optimism. The discipline that matters: only run research that could change a decision. Research that just reassures you is procrastination with a spreadsheet attached.
What deep market research is
The early Discover-stage market work tells you whether a market exists and roughly who is in it. Deep market research goes further: it puts numbers on the size, names every competitor and what they charge, and gathers hard evidence that real people will pay your price. It is the difference between “people seem to want this” and “here is how many, at what price, against whom.”
The work has four jobs. Size the addressable market honestly, not the flattering total-market figure. Map competitors in enough detail that you know exactly where you sit on price and feature. Validate the price with evidence that someone will actually hand over the money. And read the demand signal: search volume, waitlist sign-ups, pre-orders, anything that is a behaviour rather than an opinion.
The trap this activity exists to avoid
Here is the opinionated claim, and I will defend it: most deep market research is not done to make a decision. It is done to feel safe before making a decision you have already made. The two look identical from the outside. Both involve spreadsheets, competitor tables and a slide that says “the market is £2.3bn.” Only one of them earns its keep.
The test is brutal and simple. Before you start a piece of research, write down what you would do differently depending on the result. If a high number and a low number lead to the same action, you are not researching; you are reassuring. The research that changes a decision is worth weeks. The research that only changes your confidence is worth nothing, and quietly costs you the weeks anyway.
Below is the deep dive we ran on the proofing box, and the one finding in it that actually moved a decision rather than padding a slide.
Four of those rows reassured us. One of them changed what we did. That ratio is normal, and it is the whole point: the value of the dive sat in a single row, and we only found it because we went looking for the thing that could move a number.
- The flattering total-market number that no plan ever hangs off.
- A fortnight more competitor reading after the picture is already clear.
- Surveys asking “would you buy this?”, which everyone answers yes to.
- A finding that, high or low, leads to exactly the same action.
- The honest addressable number that sets the size of the first run.
- The price band a real comparable confirms or breaks.
- A demand signal that is a behaviour: a search, a pre-order, a sign-up.
- A finding you wrote a decision against before you went looking.
If a piece of research belongs in the left column, stop doing it. The discipline is knowing when to declare the picture clear enough and move, because there is no number large enough to make a launch risk-free.
How it fits the bigger picture
Deep market research sits in Stage 03 Innovate, after the earlier market and persona work and before you test with focus groups (03.07). This activity puts the numbers and the competitor map on the table; the focus groups then pressure-test the conclusions against real people rather than spreadsheets.
What it can do
It tells you how big the realistic market is, exactly who you are up against and at what price, and whether the demand signal is a behaviour or just an opinion. Done well, it sets the size of your first run and the price you can defend.
What it can’t do
It can’t make the decision for you, and it can’t remove the risk; it can only reduce it to a level you can live with. It also can’t tell you when to stop. That judgement is yours, and the cost of getting it wrong is weeks lost to research that only ever reassured.
See the full 10-stage process →
Try it yourself
Before you open a spreadsheet, list the three decisions this research is meant to inform: how big the first run is, what price you can hold, who you compete with. For each, write down what a high answer and a low answer would make you do. If both lead to the same action, cross it off. Research only the questions that survive. When the picture is clear enough to act, stop, even if it feels uncomfortable.
Want a structured starting point? Start the Free Sprint → and the GPT will frame your market questions so the numbers point at a decision.
Your deep-research checklist
Project notes: the one row that mattered
▸ From the notebook · optional reading
The deep dive on the proofing box with Dan and Anna Hartley in Stockport, and the moment we caught ourselves researching for comfort instead of for a decision.
3 min read · click to open
Dan wanted a market-size figure for the deck. Fair enough. The first pass gave us a lovely big number: every home that owns a Dutch oven, near enough. I asked the awkward question. “If that number were half what it is, would we do anything differently?” Silence. We would not. So we threw it out and sized the market we actually sell to: serious overnight-prove bakers in cold kitchens who already buy specialist kit. Far smaller, far more useful.
Where the dive earned its keep
The competitor map was where the real work was. We laid out everything from a few-pound fabric proofer to commercial proving cabinets several hundred pounds up, plus the free DIY oven-light trick that most bakers default to. The picture that mattered was the gap in the middle: nothing premium-but-domestic between the fabric bag and the catering cabinet. That empty band was the whole case for a £149 ceramic box.
We pushed on price next. Rather than ask people what they would pay, which never works, we checked what comparable specialist baking kit sold for at Lakeland and watched what the Sourdough School audience was already buying. The band held. That was a behaviour, not an opinion, and I trust behaviour.
The row that changed the plan
Here is the one finding that paid for the whole exercise. The honest addressable number, set against the empty mid-band, told us the demand was real but not vast. So instead of gambling on a big opening order, we set the first run at 500 to 1,000 units, with Year-1 around 3,000 once the signal firmed up. Every other row in the dive just confirmed what we suspected. This one row moved the money.
The temptation, once the picture was clear, was to keep reading. Another fortnight of competitor analysis, a deeper search-trend study, a survey. I called it: we knew enough to decide, and more research would only have made us feel braver, not be righter. We stopped, and we built. (Anna, to her credit, agreed before I did.)
— Innovate stage, project notes, 2026
— Next in Innovate → Test with focus groups