How to Find Competitor Market Share: 10 Sources and 3 Ways to Calculate It

Start with the part that decides everything else. A share figure is one company’s sales divided by the whole market’s sales, and almost every article on this subject treats the second number as though it were sitting somewhere waiting to be looked up. It usually is not. Whether you can find competitor market share at all comes down to a single question: is anybody measuring the total?

In some markets somebody is. Regulators count licensed operators. Retail measurement panels read till receipts. Governments count business revenue by classification code. Public buyers publish every contract they award. In those markets a real share figure exists and you can go and get it. In most business software categories nobody counts anything, and the shares circulating in press articles are estimated revenue divided by an estimated market size, which is two models stacked on each other.

1. 1Decide which share question you are actually asking. Three different numbers get called market share: share of a whole industry, share relative to your largest rival, and share of the deals you personally compete in. They need different data and support different decisions. Pick one before you collect anything.
2. 2Define the market in one written sentence. Name the buyer, the problem, the geography and the period. "Mid-market HR teams in the EU buying applicant tracking, 2026" is a market. "HR software" is not. That sentence travels with every share figure you ever publish.
3. 3Find a measured denominator, or admit there is not one. Regulators, statistics agencies, retail panels and procurement registers publish real totals. If none of them covers your market, stop hunting for a percentage of an industry and switch to a share you can actually compute.
4. 4Collect the numerator on the same basis. Whatever unit the denominator uses, revenue or units or subscribers or contracts, the competitor figure has to use it too. A revenue numerator over a unit denominator produces a number that looks perfectly respectable and means nothing.
5. 5Compute relative market share as the fallback. Divide your figure by the largest competitor's. It needs no market size at all, it is the axis the growth-share matrix was built on, and it answers the question a raw percentage usually does not: are we the leader or the challenger.
6. 6Measure your share of the deals you contest. Pull win rate by named competitor out of your own pipeline. It is first-party, measured and current, and it is the only market-share-shaped number on this page you can verify end to end.
7. 7Record the definition, the source and the date beside the number. A share figure stripped of its market definition cannot be audited, and it will be quoted back at you in a board deck a year later. Re-check quarterly, and immediately when a competitor is acquired or a new entrant appears.

“Market share” names three different calculations that get used interchangeably and should not be. They need different data, they survive different amounts of missing information, and they support different decisions. Choosing the right one is what separates a defensible number from a decorative one, and it is the step that almost never appears in the guides ranking for this query.

Absolute share, when a real denominator exists

This is the classic percentage, and it is only as good as the total underneath it. Take the total from something that counted rather than something that estimated, take the numerator on the identical basis, and carry the market definition alongside the number for the rest of its life. If the total came from a market research report, read how the report defined the segment before you quote the result: two reputable houses sizing “CRM” can differ twofold purely on scope.

Relative market share, when no denominator exists

Divide your sales by your largest competitor’s. That is it. No market total, no classification code, no report. Above 1.0 you lead; at 0.3 you are a third the size of the leader and the strategic conversation is completely different. This is the axis the growth-share matrix was built on, and it stays computable in exactly the situations where absolute share quietly becomes fiction. Its one dependency is a defensible revenue figure for the rival, which is a solvable problem: the sources for that are set out in the guide to finding competitor revenue.

Share of contested deals, which nobody else tells you to use

Filter your own closed opportunities to those where a specific competitor was in the room, then compute how many you won. That is a share of a real, countable market: the one you actually sell into. It is measured rather than modelled, it is current rather than a year old, it moves when your positioning moves, and it is the only number here you can fully audit because you own the underlying records. It also does something the other two cannot, which is tell you whether a competitor is getting harder to beat before their revenue reflects it. Win rate covers how to compute it cleanly.

1. Analyst market share trackers

IDC, Gartner and Canalys publish quantified vendor shares for the technology markets they have decided to track, usually quarterly and usually with a stated segment definition. Where one exists for your category it is the fastest credible answer on this page. Two cautions. Coverage is patchy and skewed to large markets, so a niche category is unlikely to be tracked at all. And the numbers come substantially from vendor briefings, which means a vendor who declines to brief can be under-counted or missing entirely.

2. Public filings and earnings calls

For a listed competitor, the annual report’s business section carries a discussion of competition, and management often states a share on earnings calls. Read it for three things, only one of which is the number: how they define their market, which competitors they name as real, and which segments they describe as growing. The definition and the named rivals are more durable intelligence than the share claim, because the claim is self-reported and the market boundary was chosen by the person quoting it.

3. Industry regulators and trade associations

Where an industry is licensed, somebody counts it. Banking supervisors publish deposits and lending by institution, telecoms regulators publish subscriber numbers by operator, insurance regulators publish written premiums, energy and aviation regulators publish volumes. These are measurements rather than estimates and they are usually free. If your competitors sit in a regulated sector, this is the best source on the page and most competitive research never opens it.

4. Government statistics agencies

National statistics offices publish total revenue and business counts by industry classification code. This gives you a denominator, never a numerator, and it arrives one to two years late. Used carefully it is still valuable: it bounds the market, and it exposes market-size claims that exceed the entire measured industry. Check which classification code your category actually falls under before trusting the total, because software categories are frequently spread across several.

5. Library business databases

Industry research databases such as IBISWorld, Statista and MarketLine carry a market size, a growth rate and a list of major participants with estimated shares. They are expensive commercially and free through a great many public library systems, remotely, with nothing more than a library card. That is a genuinely useful and almost never mentioned route to reports that otherwise cost four figures. Treat their share splits as informed estimates, and read the methodology note, which they do publish.

6. Retail measurement panels

In consumer goods, share is genuinely measured. Panel providers read till data from retailers and purchase diaries from households, and report unit and value share by brand weekly. It is expensive and it is the real thing. If you sell consumer products, share language means something precise in your market that it simply does not mean in business software, and it is worth being explicit about that when the two audiences read the same report.

7. Public procurement award databases

Public bodies publish what they bought, from whom and for how much. That gives you a complete, countable segment: every award in a category over a period, with vendor names and contract values. You can compute a true share of public-sector spend, which is a real measurement even though it covers only one slice of the market. In sectors with heavy public buying, that slice is large enough to be a reliable directional read on the whole.

8. Technology detection install counts

Detection services fingerprint what a public web page loads, so for any product that leaves a trace in a browser they can count detectable installs per vendor. The universe is the set of sites they crawl rather than the market, so the absolute count is not a market share. The ratio between two vendors inside the same universe, tracked over time, is a legitimate relative signal. The mechanics and limits of that detection are covered in the guide to finding a competitor’s tech stack.

9. Review site and app store counts

Review volume, category ranking and download estimates all correlate loosely with install base and are all easy to distort. A vendor running a review campaign can double its review count in a quarter without gaining a single customer. Use these to spot a vendor you had not heard of and to watch a trend line, never to state a share. Grade them low and say so when they appear in your analysis.

10. Your own CRM and contested-deal history

Every deal you have contested against a named competitor is a data point about market share that nobody else on earth holds. Aggregate them into a win rate per competitor and a trend over the last four quarters. This is the source that turns a research exercise into something the sales team recognises, and it is sitting in a system you already pay for. Keeping the competitor field populated on every opportunity is the only work involved.

Both are legitimate, they answer different questions, and quoting the flattering one without saying which it is has been a marketing habit for decades. Unit share divides units sold by total units. Revenue share divides sales value by total market value. They agree only when everyone charges the same price, which happens in no real market.

There is a useful trick hiding in the pair. Divide revenue share by unit share and you get a relative price index: above 1.0 means the company sells above the market average price, below means it sells beneath it. A competitor whose unit share is climbing while their index falls is buying share with discounting, which is visible in the arithmetic long before it appears in anyone’s commentary.

1. 1Find the denominator first. Ask what total the percentage was taken against and where that total came from. A share figure whose denominator nobody can name is an opinion with a decimal point.
2. 2Classify it as measured, estimated or claimed. Regulator and panel figures are measured. Analyst tracker figures are estimated. Figures in press coverage are usually a company’s own claim reprinted. All three get called market share and only the first is a fact.
3. 3Check the shares sum sensibly. Add up every published share for the market. Well over 100% means the sources used incompatible market definitions; far under it means a large unattributed remainder that may itself be the real competitor.
4. 4Sanity-check against headcount. A vendor claiming to lead a market with a fifth of the leader’s employees is claiming a productivity difference that would be famous. Cross-checking against competitor headcount catches inflated claims quickly.
5. 5Test it against your own pipeline. If a competitor supposedly holds a third of your market but appears in one deal in twenty, either the market definition does not describe the market you sell into, or the figure is wrong. Both are worth knowing.

Market share research carries a specific legal risk the other competitive data points do not, and it catches people out because the risky activity looks like ordinary industry collaboration.

• Do not exchange commercially sensitive data directly with competitors. Sharing current or forward-looking volumes, prices or capacity with a rival, including through a trade association working group, is treated as unlawful information exchange by competition authorities in both the EU and the US, and the exchange itself can be the infringement without any price ever being agreed. Benchmarking compiled by an independent third party from aggregated, historical, anonymised inputs is the accepted form, and the distinction is not a technicality.
• Respect the licence on a research report. Analyst reports are licensed to named seats. Circulating the PDF, republishing its charts or posting a share figure with the chart attached generally breaches that licence. Quoting a number with attribution is normally permitted; check the terms, because they differ by house.
• Do not misrepresent who you are. Requesting a vendor briefing, an analyst inquiry or a market report while posing as a customer or a journalist is deceptive, breaches the professional standards competitive intelligence practitioners work to, and is the fastest way to make a legitimate research function look like something else.
• Read published figures as much as you like. Filings, regulator data, statistics releases, procurement registers and your own records are all fair game, and assembling them into a share estimate is exactly what the discipline is for.

• A precise share in an unmeasured market. Nobody counts most business software categories, so no correct answer exists to be found. Proxy: relative share against the leader, plus share of a countable universe such as public contracts or detectable installs, with the universe named.
• Share by segment or by geography. Published shares are nearly always global and whole-market, while the decisions you make are segment-level. Proxy: your own win rate against that competitor filtered to the segment, which is segment-level by construction.
• Share of wallet inside an account. How much of a shared customer’s budget goes to them rather than you is known only to the customer. Proxy: ask in account reviews, and watch which of you the customer sends to their integration and procurement conversations.
• Share momentum in real time. Every published share figure describes a period that has already closed, typically one to four quarters ago. Proxy: leading indicators that move first, namely new customer logos, hiring by function, and your own contested win rate.
• The share of privately held new entrants. The competitor most likely to take share from you next is too small to appear in any tracker. Proxy: mentions in your lost-deal reasons, which is where a new entrant surfaces first by a wide margin.

Market share ages in two different ways and they need different cadences. The published percentage moves slowly, so re-checking a tracker or a regulator release quarterly is enough. The things that change share move fast: a competitor announcing a large customer, entering a new country, launching into an adjacent category or closing a funding round. Those deserve to reach you the week they happen, not in a report the following spring.

Keep the share figure, its market definition, its source and its verification date together in one place, and note the events that would force a re-read. A competitor tracking spreadsheet is enough structure for a handful of competitors. Flares watches the share-moving events continuously, which matters here because by the time a market share number is published, the movement that produced it happened three quarters ago.