Competitor Matrix: The Weighting Is the Whole Exercise

A competitor matrix with unweighted rows is opinion wearing a spreadsheet. You list twelve criteria, score everyone out of ten, add up the columns, and the tool you already liked wins by four points. Nobody notices that “SOC 2 certification” and “has a dark mode” just counted the same.

The grid is not the hard part. Any competitor matrix can be drawn in ten minutes. The hard part is deciding which three criteria actually decide your deals, and being willing to say the other nine matter less. That decision is the analysis. Everything else is formatting.

What a competitor matrix is actually for

There are two shapes people mean by this, and they answer different questions.

A weighted scorecard ranks a set of rivals against criteria you have deliberately weighted. It answers “who is ahead, and on what.” It is the version that belongs in a board deck or a roadmap argument, because it forces you to commit to what matters before you score anyone.

A 2x2 landscape map plots the same competitors on two axes. It answers “how does this market cluster.” It is worse at ranking and much better at showing you the empty quadrant nobody is serving, which is usually the more interesting finding.

You want both, built from the same numbers. The scorecard tells you where you stand; the map tells you where you could stand instead. Building them separately is how teams end up with two artifacts that quietly disagree.

Step one: cut the criteria list in half, then cut it again

The instinct is to be thorough. Thoroughness is what ruins the exercise.

Every row you add dilutes the rows that matter. A twelve-row grid where three rows decide deals gives those three rows 25% of the weight between them. Your matrix will then rank a competitor as “close” when they are in fact losing every deal they enter against you, because they win on eight things buyers never asked about.

The filter is a single question: has a buyer ever changed their mind because of this? Not “would a buyer care,” but would they switch over it. Support responsiveness usually survives that question. Number of integrations usually does not; what survives is the two integrations your buyers actually name.

Six to eight rows is a healthy grid. If you have fifteen, you have a feature inventory, and a feature inventory is a different document with a different job.

Step two: weight before you score, always

Weight the criteria before you look at a single competitor. This ordering is not fussiness. It is the only defense against the thing that makes most matrices worthless.

If you score first, you will weight to fit the scores. Everyone does. You will notice you are losing on price, decide price is “not really how this market buys,” drop its weight to a two, and produce a document that proves what you walked in believing. Weighting first makes that move visible, because you have to argue the weight down in front of people who saw you set it.

Use a 1 to 5 scale and expect discomfort. If four of your six criteria are fives, you have not weighted anything. Two decisive criteria, two important ones, and the rest as context is a realistic distribution for most B2B SaaS categories.

Then the math is just:

Weighted score = Σ (score × weight) ÷ Σ (weight)

Dividing by total weight is what keeps columns comparable when you inevitably leave some cells blank.

Step three: score from evidence, and leave blanks blank

Score from artifacts you can link to. Pricing pages, changelogs, review profiles, docs sites. A score you cannot source is a memory, and memories about competitors skew toward whatever lost you a deal most recently.

Where you do not know, leave the cell empty. Do not guess a five. A guessed midpoint is worse than a blank, because a blank is honest about your coverage and a five silently drags a rival toward the middle of the pack. In a properly built scorecard, blanks are excluded from the average rather than counted as zero, so a competitor scored on four of six criteria still ranks fairly against one scored on all six.

The blanks are also a to-do list. A column with four empty cells is telling you that you are arguing about a competitor you have not actually researched.

Step four: read the map, not just the total

The weighted total gives you a rank. The rank is the least useful output.

What you want is the shape. Plot two criteria as axes and you will usually see one of three things: everyone clustered in one quadrant, which means the category is undifferentiated and positioning is available cheaply; you alone in a quadrant, which means you either found a wedge or a market nobody wants; or a rival drifting toward you, which is the finding that should change your roadmap this quarter.

None of that is visible in a column of totals. The number 7.4 does not tell you that two competitors just moved into your quadrant.

How to use the output, by role

For founders: the matrix is an input to one decision: where the next two engineering months go. If the output does not reorder your roadmap or change a price, you built it for reassurance.

For product marketing: the two criteria with the biggest weighted gap in your favor are your messaging. Not the ones you find most interesting. The whole exercise exists partly to stop teams from leading with the feature they are proudest of instead of the one buyers weigh most.

For sales: the rows where a specific rival beats you are next quarter’s objection handling. Feed them straight into a battlecard rather than letting each rep improvise a different answer.

Where matrices rot

A competitor matrix is accurate the week you build it. Then a rival ships the feature you scored them a three on, changes their entry price, or picks up 200 reviews, and the grid keeps saying what it said in September.

This is the failure mode worth designing against, because a stale matrix is more dangerous than no matrix. Nobody makes a decision from a document they know is missing; plenty of teams make decisions from a confident-looking grid that is eight months old. Date the thing. Re-score it quarterly at minimum, and immediately after any competitor pricing change, since those tend to invalidate the highest-weighted row you have.

If you want to see how fast that happens, look at how much movement there is in a single category’s pricing over a year: our Klue pricing breakdown exists because the “contact sales” answer keeps changing underneath everyone’s comparison docs.

The one move worth making this week

Take the competitor comparison your team already has, the one in a Google Doc or a slide, and add a weight column. Do not re-research anything. Just assign 1 to 5 to each existing row, in a meeting, out loud, before anyone looks at the scores.

Most teams find at least one row they have been treating as decisive that nobody can defend above a two, and one row scored casually that turns out to be a five. That reordering, on its own, is usually worth more than the next round of research.

You can do it in the browser with our free Competitive Matrix Builder: weight the criteria, score up to four rivals, and flip to the 2x2 view to see the landscape the totals hide. It runs entirely client-side and exports to CSV, PNG or PDF.

Outmano watches your competitors’ pricing, features, content and reviews continuously, so the scores behind a matrix like this stay current instead of quietly expiring. It is the difference between a grid you trust and a grid you rebuild from scratch every quarter. See how it works.

Frequently Asked Questions

What is a competitor matrix?

A competitor matrix is a grid that scores your product and its rivals against the same criteria, so a comparison that normally lives in opinions becomes something you can check. It comes in two forms: a weighted scorecard that ranks competitors on criteria you have weighted by importance, and a 2x2 map that plots them on two axes to show how the market clusters.

How many competitors should a matrix include?

Three to five columns, including your own. Two gives you no sense of the category; more than five turns scoring into data entry and you will start guessing cells to finish the exercise. Pick the rival you lose to most, the fastest-growing one, and one or two adjacent threats.

What is the difference between a competitor matrix and a SWOT analysis?

A SWOT analysis is qualitative and usually about one company at a time, covering its strengths, weaknesses, opportunities and threats in prose. A competitor matrix is comparative and numeric: the same criteria applied across several rivals so you can rank them. Use SWOT to think about one competitor deeply, and a matrix to see several at once.

How do you weight criteria in a competitive matrix?

Assign 1 to 5 to each criterion before scoring anyone, based on whether a buyer has ever changed their decision over it. Expect no more than two or three fives. Weighting after you score is how teams accidentally reverse-engineer the conclusion they already held.

How often should you update a competitor matrix?

Quarterly at minimum, and immediately after any competitor pricing change, since price is usually one of the highest-weighted rows. Anything built on features and pricing starts decaying the week you finish it, which is why continuous tracking beats a calendar reminder.

Can a competitor matrix include your own product?

Yes, and it should. Scoring yourself with the same criteria and the same evidence standard is the only way the totals mean anything, and it is where teams discover they have been generous to themselves on onboarding or support. If you cannot bring yourself to score your own product honestly, the matrix will not survive contact with a win/loss review.