Most founders carry their market in their head — a few rivals they watch, a vague sense of where the wind is blowing, and a slide deck that's six months stale. Then the board asks "who else does this and why do we win?" and you improvise, or you set a roadmap against a competitor that pivoted last quarter. This system spends 45 minutes turning your scattered observations, a few competitor pages, and some analyst notes into a single market-landscape.md: the real and adjacent competitors (including "do nothing" and "a spreadsheet"), where each wins and loses, the few market shifts that actually move your bets, the positioning gap you could own, and a short watch-list. It's a conversation in Claude Desktop — open the app, drag your files into the file pane, ask in plain language. Build it once, refresh it quarterly, and your quarterly planning and board narrative both start from a real read instead of a guess.
- Raw competitor material — drop their landing pages, pricing pages, a few recent posts, or
g2-reviews.txtinto amarket/folder and open it in Claude Desktop. Public material only. - Any analyst notes, market reports, or category write-ups you've collected — paste the text or save them as files alongside the competitor pages.
- Your own scattered observations — the half-formed notes in your head about who you lose deals to and why. A few bullets in a
notes.mdis enough; Claude will structure them. - Your one-line positioning and your two or three real strengths, so Claude can find gaps relative to you — not in the abstract.
-
Dump the raw material and have Claude synthesize the landscape
Don't try to write the landscape from memory — you'll only list the rivals you already worry about and miss the adjacent ones. Put everything in the
market/folder, open it in Claude Desktop, and let Claude read across all of it at once. When it asks to read the folder, approve it in the "Ask permissions" prompt; nothing is written yet, so there's no diff to review here.you askIn the market/ folder are competitor pages, some analyst notes, and my own rough observations. Read across all of it and map the landscape for me: who are the direct competitors, who are the adjacent or indirect ones, and — importantly — include the 'non-consumption' options like doing nothing, a manual process, or a spreadsheet. For each, one line on who they're for and what they actually do. Don't rank them yet, just lay out the field.what you get back A structured map of the whole field — direct rivals, adjacent players, and the "a spreadsheet / do nothing" options you'd have forgotten — each with a plain one-liner. The picture you couldn't hold in your head, assembled from your own pile.
Including "do nothing" and "a spreadsheet" as competitors is the move most founders skip. For early products that's who you actually lose to — name them so your strategy answers them.
-
Pressure-test it: where do we actually win and lose
An honest landscape names your weaknesses, not just theirs. Push Claude past the marketing-page version of each competitor — "strong" and "loud" are different things — and make it say where you genuinely lose, so the brief is something you can plan against rather than a comfort blanket.
you askHere's our positioning and our real strengths: [paste]. Go competitor by competitor and tell me honestly: where do we genuinely win, where do we genuinely lose, and — separately — where is each competitor actually strong versus just loud and well-marketed? Be specific and cite which page or note you're drawing each claim from. Don't flatter us.what you get back A frank win/lose read per competitor, with the "strong vs just loud" distinction called out and each claim tied back to a source — including the uncomfortable places you lose, which are the ones worth knowing.
If the read feels too kind, say so: "that's the generous version — where would a skeptical customer say we're behind?" The harsh version is the useful one.
-
Find the positioning gap you could own
The payoff of a landscape isn't the list — it's the wedge: the position nobody owns that actually fits your strengths. Cross the gaps in the field against what's true for you, so the angle is both differentiated and defensible, not just contrarian. Claude proposes; the wedge is your call.
you askLooking at the whole field and where everyone clusters, where are the open positioning gaps — angles, audiences, or use-cases nobody clearly owns? Give me 3 candidate wedges. For each, say which of our real strengths it leans on, why it'd be hard for the others to copy, and what we'd have to be true to claim it credibly.what you get back Three candidate wedges, each tied to a real strength of yours and a real gap in the field, with an honest "what you'd have to back up" attached — a shortlist to choose from, not a decision handed to you.
Claude proposes wedges; you pick. The one you can defend to a customer and a board — and actually deliver — is the wedge, even if it's not the cleverest on the list.
-
Assemble the reusable market-landscape.md
Package everything into one artifact you'll reuse for quarters: a summary, a competitor table, the few market shifts that matter, your chosen wedge, and a watch-list. Claude writes this as a real file — when it does, you'll see an accept/reject diff in Claude Desktop. Read it before you accept; this is the document your roadmap and board narrative will lean on.
you askAssemble all of this into a single market-landscape.md: a one-paragraph summary of the market, a competitor table (each row: name, who-they're-for, where-they-win, where-they-lose, strong-or-just-loud), the 3-4 market shifts that actually affect our bets, our chosen wedge with its rationale, and a short 'what to watch' list of signals that would change this picture. Keep it tight and skimmable. Then, separately, give me 3 concrete implications this landscape has for our roadmap.what you get back A clean, paste-ready
market-landscape.mdplus three roadmap implications — "the spreadsheet is our real competitor in segment X, so onboarding speed matters more than feature breadth," the kind of line that turns a market read into a planning decision.Save this file in the
market/folder. It becomes the input your Run a quarterly planning cycle and Write the board narrative playbooks start from — so they begin with a real read, not a fresh guess.
- Quarterly refresh: keep the
market/folder and re-run this every quarter. Drop in the new pages and notes, then ask Claude to diff against last quarter's `market-landscape.md` — what shifted, who moved, what's new — so you track the market's motion instead of re-snapshotting from zero each time. The refresh takes under an hour. - Investor / fundraising lens: before a raise, ask Claude to reframe the same landscape for the "why now and why us" story — which market shift makes the timing right, and how the wedge reads to an investor versus a customer. Same source material, a sharper narrative cut.
- Adapt, don't translate, for the Gulf market (Arabic): if you operate in the UAE or wider MENA-Gulf, the global landscape is rarely the local one — the relevant competitors, the "do nothing" baseline, the regulatory and procurement realities, and who the buyer trusts can all differ. Build a parallel Arabic
market-landscape-ar.mdreasoned from local sources, not a translation of the English file. A position that's wide open globally may already be crowded locally, and vice versa. - Save the move (Power Track): once the prompt sequence works, advanced users can store it as a
/landscapecustom command (see the Features tab) so each quarter's refresh is one command over the updated folder. On Desktop you can keep the dialed-in prompts in a note and paste them — the custom command is the opt-in terminal twin, not a requirement.
- Verify before you act or quote. Claude will state competitor facts, funding figures, pricing, and feature claims with total confidence — and it can be wrong or out of date, because it only knows what's in your files (and its training), not the live web. Treat the brief as a sharp draft, not a source of truth: verify anything you'll put in a board deck, a roadmap decision, or a public claim against a real, current source. A confident sentence is not a checked one.
- Never feed it confidential material. Use public pages, published reviews, and your own observations. Don't paste anything under an NDA, a competitor's leaked or confidential document, or material from a paywalled tool you agreed not to share. Competitive research that crosses that line is a legal exposure, not an edge — and it taints every decision built on it.
- The wedge is your call, not Claude's. Claude can surface candidate positions and pressure-test them, but choosing what your company will stake its roadmap on is a strategic decision only you can defend to customers, your team, and the board. Use Claude's three wedges as a shortlist to react to, not a verdict to accept.
- Don't let the landscape replace your customers. This is one input. What your actual buyers say in calls and lose-reasons should outweigh what a rival's landing page claims — a beautiful competitive map built only from marketing copy can point you confidently in the wrong direction.
you'll end up with A reusable `market-landscape.md` — summary, competitor table, the market shifts that matter, your chosen wedge, and a watch-list — that grounds your roadmap and board decisions in a real read instead of a guess, and refreshes in under an hour each quarter.
Questions people ask
- Does Claude research my competitors on the live web?
- No. Claude works from the files and notes you put in the `market/` folder plus its training data — it isn't browsing competitor sites in real time. The brief is exactly as current as the material you gather, so collect fresh pages and date them, and verify any fact, figure, or feature claim against a real source before you act on it.
- What do I need to gather before I start?
- Public competitor material (landing and pricing pages, recent posts, published reviews), any analyst or market notes you have, your own rough observations about who you win and lose against, and your one-line positioning plus two or three real strengths. A folder of files and a few bullets is enough — Claude does the structuring.
- Who decides the positioning wedge — me or Claude?
- You do. Claude proposes candidate wedges and pressure-tests each against your strengths and the gaps in the field, but choosing the position your roadmap will stake on is a strategic call only you can defend to customers and the board. Treat its three options as a shortlist to react to, not a decision to accept.
- How often should I refresh the market-landscape.md?
- Quarterly works well. Keep the `market/` folder, drop in the new material, and ask Claude to diff against last quarter's file so you see what shifted rather than re-building from scratch. The refresh takes under an hour, and the file feeds straight into your quarterly planning and board narrative.