How to Build a Competitor Landscape in One Day With a Source for Every Claim

To build a competitor landscape in a day, define the exact market question first, list every company a buyer would realistically compare you with, sort them into segments, profile each one from public sources, then plot them on two axes that matter to buyers. Record a source and a date for every claim.

That is the whole method. The rest of this guide is the detail that turns it from a rough sketch into something you can put in front of a board, an investor or a sales team without anyone asking where a number came from. If you would rather skip the manual work, our AI market research agent runs the same steps and keeps the result current, but the method below works with a browser, a spreadsheet and a focused day.

Why a competitor landscape goes stale so fast

Most landscapes are built once for a deck and then quietly rot. A competitor changes its pricing page, another raises a round and starts hiring a sales team in your region, a third pivots its messaging toward your core segment. None of that shows up in a slide made three months ago.

The second problem is trust. A landscape without sources is an opinion. When a board member asks "how do we know they charge per seat?" and the honest answer is "someone told me", the whole document loses weight. The fix is cheap: every row gets a link and a capture date. It adds perhaps ten seconds per fact and it changes how the document is received.

Step one, write the question before you open a browser

A landscape answers a question. "Who are our competitors?" is too vague, because the answer depends on who is buying and why. Better questions look like these:

• Which tools would a 50 person US software company shortlist when replacing spreadsheets for expense approval?
• Which vendors sell payroll software to small businesses in Germany, and how do they price it?
• Who competes for the budget of a VP of product marketing who wants competitor alerts?

Each question fixes three things: the buyer, the job they are hiring a product for, and the geography. Write yours at the top of the document. When you are unsure whether a company belongs on the list, go back to the question and ask whether that buyer would actually compare it with you.

Step two, build the long list in about two hours

Aim for 20 to 40 names on the first pass. You will cut some later, but it is much easier to remove a company than to discover in a sales call that you missed one. Good public sources for the long list:

• Search results for buyer phrases. Search the words your buyer would type, such as "expense approval software for startups", and note every vendor on the first two pages, including ads. Paid placements tell you who is spending to win that buyer.
• Review and comparison sites. Category pages on software review sites list vendors buyers actually evaluate. Note the category names too, because they show how buyers describe the problem.
• "Alternatives to" pages. Search "[known competitor] alternatives". Competitors write these pages about each other, which surfaces names you would not think of.
• Funding and company databases. Public funding announcements and company registers reveal newer entrants that do not rank in search yet.
• Your own pipeline. Ask sales which names come up in lost deals. Ask support which tools new customers moved from.

Put every name in a sheet with one column for the source where you found it. That column becomes surprisingly useful later, because it shows which channels your competitors use to reach buyers.

Step three, segment before you profile

Profiling 40 companies one by one without structure takes days. Segmenting first cuts the work, because companies in the same segment share most attributes and you only need to note the differences.

Typical segment lines in B2B software:

• Target customer size. Self-serve for small teams, sales-led for mid-market, enterprise suites with long contracts.
• Breadth. Point solution that does one job versus a platform that bundles several.
• Delivery model. Pure software, software plus services, or agencies and consultancies that solve the same problem by hand.
• Adjacent substitutes. Spreadsheets, generic tools and internal builds. They rarely appear in market reports, but they are often the real competitor.

Three to six segments is usually right. Fewer and the map says nothing, more and nobody remembers it. Our market map software groups companies this way automatically and shows the positioning grid next to the list, which is handy when you want to test different axes.

Step four, profile each competitor from public sources

Use the same fields for every company so the rows are comparable. A practical set:

Two rules keep the profile honest. First, quote rather than summarize when you capture positioning, because your summary drifts toward what you expect. Second, write "not public" when something is not public. Guessing a quote-only vendor's price and presenting it as fact is how landscapes lose credibility. If you include third party estimates, label them as estimates and link the source.

Step five, choose axes buyers actually care about

The classic two by two plot is useful only when the axes reflect real buying decisions. "Innovation" versus "vision" means nothing to a buyer. Axes that usually work:

• Self-serve versus sales-led buying
• Point solution versus broad platform
• Price per year, low to high
• Time to first value, minutes versus months of setup
• Small team focus versus enterprise focus

Pick the two axes that best explain why a buyer chooses one vendor over another in your question. Plot each company using the evidence in its profile, not gut feeling. If you cannot place a company on an axis from public evidence, that is a finding in itself: the vendor is not communicating it clearly.

Step six, find the gaps and write them down

The map is a means, not the deliverable. The deliverable is three to five short findings, each tied to evidence. For example:

• "No vendor in the self-serve segment offers approval workflows for more than one entity. Source: pricing and feature pages captured on the dates in rows 4 to 11."
• "Every enterprise vendor gates pricing behind a sales call. Buyers who want to start this week have no option above the small team tier."
• "Two competitors added the same integration in the last quarter, according to their changelogs. This is becoming table stakes."

Findings like these give strategy, product and sales something to act on. A recommendation follows naturally, and you can track whether it holds as the market moves. Our strategy module turns findings into recommended moves and battlecards, but a plain list in the document works too.

A realistic one day schedule

That is a full day for one person, and it is tight. The profiling step is where most of the time goes, which is why it is the step worth automating first.

Keeping the landscape current after day one

A landscape is accurate on the day you build it. To keep it useful, watch the sources that change most often:

• Pricing pages. Changes here are the clearest signal of strategy shifts. See our guide on how to monitor competitor pricing changes.
• Homepage headlines. A new headline often means a new target buyer.
• Changelogs and release notes. They show what each competitor is investing in.
• Careers pages. Hiring an enterprise sales team in a new region usually comes before a market entry announcement.

Checking all of that by hand every week across 30 competitors is the point where most teams give up. That is the job our competitor monitoring does in the background, with competitor alerts when something material changes and a weekly brief for the rest.

Common mistakes to avoid

• Listing only the companies you already know. The competitor that hurts you is often one you have not heard of yet. The long list step exists to catch it.
• Mixing facts and opinions in the same cell. Keep a separate "our take" column so readers can tell evidence from interpretation.
• Ignoring substitutes. If buyers solve the problem with a spreadsheet today, the spreadsheet is on the map.
• No dates. A price captured last year and a price captured yesterday look identical in a table unless you record when.
• Too many axes. One clear grid beats four confusing ones.

Frequently asked questions

How many competitors should a landscape include?

Start with 20 to 40 on the long list and keep everyone a real buyer would compare you with. Most finished landscapes show 10 to 30 companies. Fewer than ten usually means substitutes or newer entrants were missed.

What is the difference between a competitor landscape and a market map?

A competitor landscape focuses on the companies you compete with directly and how they position themselves. A market map is wider: it includes adjacent segments, substitutes and often a market size estimate. Many teams build the landscape first and extend it into a map. Our guide to TAM, SAM and SOM covers the sizing part.

Is it legal to collect competitor information this way?

Collecting information that companies publish openly, such as pricing pages, product pages, press releases and job listings, is standard practice. Do not misrepresent yourself to obtain non-public information, do not use leaked documents, and respect the terms of the sites you use.

How often should the landscape be updated?

Review pricing and positioning at least monthly, and refresh the full landscape every quarter or before any major decision. Weekly monitoring of pricing pages and changelogs catches the changes that matter between full refreshes.

Can AI build a competitor landscape on its own?

AI can do most of the collection and first draft, which is the slow part. The output is only trustworthy when every claim carries a source you can open, so check that any tool you use links each fact to the page it came from. Every fact in a MarketAgent report has a source and a capture date. You can see what that looks like on the pricing page and in the free one-time demo on the homepage.