You've built a design tool that runs in a browser. Should you take it into a market two incumbents already own?
Three questions decide it. Each one below shows what its answer says, how well that answer is stood up, and the factors underneath that produce both — including the ones nobody has got to yet. Then the decision, and what it committed to.
Illustrative — invented figures; several tools are still being built.
- The need
1. Is the need real?
ReadsClearly realone factor judged
The obvious answer: “Designers want a faster, cleaner canvas. Build a better one and they'll switch.”
- Unmet NeedsAcute & validatedLaunch Ready31 people interviewed across 14 teams — and only 11 of them were designers. The top unmet need belongs to the engineer and the reviewer: seeing the current design without asking anyone to export it. Asked directly about the drawing tools, twelve of fourteen designers said what they already have is fast enough. Two of three teams observed kept a shared drive of dated PNG exports, refreshed by hand before each review.Needs Map
The obvious need came back served. The need that came back starving belongs to somebody nobody was selling to. You are not selling a better canvas. You are selling the file. Which makes the buying unit the team rather than the seat — and that is a different market, not a bigger slice of the same one.
- The market
2. Is the market worth it?
ReadsMixedmean 5.5 of 3 judged
The obvious answer: “Everybody knows roughly how many designers there are. Multiply and you have the market.”
- Market SizeLargeLaunch Ready$738M top-down from 4.1M professional design seats at a designer-seat price. $1.16BN bottom-up from 145,000 software teams priced on whole-team access. The two are 44% apart, and the gap is ① arriving through the sizing: count seats and you count designers; count teams and you count everyone who needs the file.Market Model
- Market GrowthWith the economyLaunch ReadyDesign-tool spend is growing a little faster than software budgets generally, but not dramatically — most of the increase is teams adding non-designer seats to files they already pay for, not a new category of buyer showing up.Market Model
- FragmentationFew or many buyersLaunch ReadyTwo incumbents hold most of the paid seats. Concentrated rather than fragmented, which scores low here for the reason it always does — share has to be taken off somebody rather than assembled.Value Chain
- Regulatory RiskNobody has judged thisNot looked atMarked not relevant to this project. Nothing regulates a design file — no jurisdiction gates the product, and no privacy or accessibility regime bears on the decision to enter, because those apply to what customers build rather than to the tool. It is out of ②'s answer rather than scored favourably and averaged in.Regulatory Review (in build)
Two honest methods 44% apart is not a failure to converge. The market is a different shape than the seat count assumes. The entry case is built on the bottom-up read, and the spread stays on the page rather than being averaged into one number nobody can defend. Regulatory Risk is ruled out here rather than scored 10 — a factor that does not apply is not a favorable one, and averaging it in would have flattered this question by a whole level.
- The edge
3. Can we win it?
ReadsChallengingmean 6.6 of 5 judged
The obvious answer: “It's obviously better. People will switch.”
- Competitive IntensityIntenseLaunch ReadyTwo entrenched rivals with years of file-format lock-in and an installed base that opens on a double-click.Competitive Landscape
- Substitution ThreatModerateLaunch ReadyThe substitute is the tool they already pay for, already know, and — the part that decides it — already have approved.Competitive Landscape
- DifferentiationUniquely positionedLaunch Ready“Everyone who needs the file is in the file” has been tested on buyers and on the people who are not buyers. It lands harder on the second group, which is the whole thesis.Value Proposition
- Pricing PowerAdequateLaunch ReadyCheaper per designer seat, and that is not where the money is. The real saving is about a day per feature of spec-writing — which appears on nobody's budget line and so cannot be invoiced against.Price Case
- Qualification & Switching BarriersNeutralLaunch ReadyGetting onto an IT-approved software list is the barrier. Nobody has costed how long it takes, or what it is worth once you are through it.Qualification Path (in build)
The value is real and lands in a department that is not the one holding the budget. You win the team and lose the procurement form — unless the tool arrives before procurement is asked anything. Which is what the free tier is for. It is not a discount and it is not pricing: it is the route in, and it is the decision doing ③'s work.
◆ The decision
The three read Conditional Go — Selective / Niche — ceiling Enter small.
Answered on nine of ten — five of six on market attractiveness, four of four on ability to win. Open: Regulatory Risk.
Launch◇ Free to view, paid per editing seatAbove the ceiling.
A decision, not a rating — Launch · Enter small · Reshape the entry · Wait / fast-follow · Don’t enter.
The obvious answer: “Undercut them on price and run a free trial.”
Free for the people who were never the buyer is what turns ①'s finding into a way in. It removes ③'s approved-list objection by changing who has to ask: the tool arrives through the people who need the file, and procurement meets it once it is already in use. Three routes were costed and this is the only one where ①'s finding does any work.
After the decision — the build
recorded against the lines you already run, not scored — and the assumption is written down
The obvious answer: “It's software. We'll build it.”
- DevelopmentNew productApprovedA custom rendering engine, not a framework choice — professional-grade vectors at 60fps in a browser. Signed off as a multi-year build with the money attached: a cost, not a blocker.
- Production readinessNew productGood as isIt ships as a URL. There is no release process to build and no version to support in the field.
- CapacityGood as isServing more teams is a server bill, not a plant. Nothing here caps the ramp.
The hard part is the one you can buy with time. The build is not clear here and does not need to be — an approved investment is a decided cost, not a caution. What matters is that it was priced before the decision rather than arriving after it.
What actually happened here
The market was owned, the incumbents were good, and the drawing tools were never the problem. The decision turned on who was locked out of the file. Once ① said the unmet need belonged to the engineer and the reviewer, ② stopped being a seat count and the decision stopped being a price. The free tier is not a discount — it is the distribution.
It is not a clean sweep either. ② reads Mixed because a market held by two incumbents is a market you take share from rather than assemble, and ③ reads Challenging because the substitute is already installed and already approved. Each also carries a factor at 4, so neither could read higher whatever its mean did — the min rule, refusing to let strong answers bury the one that decides the sale.
Once the decision is recorded, the project stops being a set of questions and becomes a place: the build and what it costs, actuals against the plan, and the triggers you set before the data arrived. It recurs rather than completes, which is why none of it is a question. See the method for where it sits.
Now do this on the decision you actually have.
Same three questions, your product and your market — the need evidenced, the size defended, the value priced — and a decision you can show the working for. The tools are open on a trial while the product is still in development.