What a competitor SWOT analysis is
A competitor SWOT analysis sorts one rival into four boxes. Strengths and weaknesses are inside the company: what it has built, what it does badly. Opportunities and threats are outside it: the market, the platforms, the buyers. All four boxes describe the same company, and each line should be a fact someone can check.
Why most SWOTs are useless
Most SWOTs are written in a meeting. Someone says "strong brand", someone else says "expensive", and both go in a box. Nobody writes down how they know, so nobody can tell next quarter whether anything changed.
The same grid tends to come back every year. "Strong brand" is always a strength. "New competitors" is always a threat. A line that would be true of any company says nothing about this one.
We took the SWOT out of our own report in August, for three reasons. The boxes had two subjects: strengths and weaknesses were about the competitor, opportunities and threats were about our customer, and a reader scanning four identical boxes reads them as four parts of one thing.
Every line was a copy of a number that already sat in another section, with its source and link, and in the SWOT it had neither. And the matrix built on top of it produced sentences like "attack the category where they are weakest", which nobody can act on.
The grid was never the problem. The evidence behind each line was. So this is the version we would stand behind: one subject in all four boxes, every line a number, every number dated.
How it works: four boxes, one subject
The note on writing a competitor study says a SWOT is about you, with rivals as one of four boxes. This turns the grid around. The rival is the subject of every box, and your own company appears nowhere until the next step.
| Box | Inside or outside the rival | What counts as evidence |
|---|---|---|
| Strengths | Inside | Things they have built that show up in public numbers: reach, catalogue, pages, policies. |
| Weaknesses | Inside | Where their own product or service lets buyers down: ratings, complaint themes, the worst-rated products. |
| Opportunities | Outside, working for them | Conditions in the market or on platforms they have not used yet, and what buyers are not asking them to change. |
| Threats | Outside, working against them | Things they do not control that are moving the wrong way: how others describe them, where the public score is heading. |
One test keeps each line honest. Could you open the source and see the number? If not, it is an opinion, and it goes in a separate list of things to check.
Worked on Loop Earplugs
Loop sells earplugs direct to consumers. Every line below comes from Loop's report, most of it measured on 28 September 2026. The complaint counts and product ratings come from the reading on 24 September, and the ad counts from the ad library as last collected on 6 September.
| Strengths (inside) | Weaknesses (inside) |
|---|---|
| Named in 79% of AI answers across 18 questions and five engines, and named first in 69%. 4,804 landing pages indexed, out of 9,996 URLs. A 100-day return window and a 24-month warranty. 6,989 live ads, and only 6% carry a discount. |
Trustpilot 3.6 on 8,133 reviews, against 4.35 on their own store. The newest reviews average 2.58. Of 282 low-rated reviews out of 2,500 read: fit and seal 30, not enough blocking 25, comfort and pain 23. Their two lowest-rated products are bundles: Good Night and Travel, both at 3.8. |
| Opportunities (outside, for them) | Threats (outside, against them) |
| Price and value is only 10 of the 282 complaints. Buyers are not asking them to be cheaper. The engines disagree: Loop is in 89% of Perplexity's answers and 67% of Copilot's. The weaker engine is ground they have not taken. |
Being named is not being recommended: only 37% of the AI answers that mention Loop frame them positively. 48% of the newest Trustpilot reviews are one star, and 108 new reviews arrived in 30 days. The public score is still being written, and not in their favour. |
The full set of measures behind the grid sits on Loop's benchmark. Each line above is a row there, with its source.
What the grid says when every box is a number
Loop is strong wherever a buyer looks before buying, and weak where the buyer ends up after wearing the product. The strengths are all reach: AI answers, thousands of landing pages, thousands of ads, a return policy that lowers the risk of trying. The weaknesses are all about the ear: fit, blocking, comfort, and two bundles that score worst in the range.
That is a pattern a meeting would not produce. A meeting would have written "strong brand" and "premium price". The numbers say price is barely a complaint, and that the brand's reach is running ahead of how the product feels.
The threats follow from the weaknesses. The complaints do not stay on Loop's store. They surface on Trustpilot, where the newest reviews run a full point below the average, and in AI answers, where most mentions carry no positive framing. The weakness is internal. Where it gets read is not.
The opportunities are thinner, and that is honest. Public data on one company mostly describes that company. Two lines are what the data supports, and neither is a guess.
How to make one
- Pick one rival and one date. A SWOT of three competitors at once turns into averages. Write the date at the top.
- List every public number you have on them: prices, ratings, reviews, ads, pages, AI answers, traffic. Note the source and the day of each.
- Sort each number into a box by asking two questions. Is it inside the company or outside it? Does it help or hurt them?
- Cut any line that would be true of most companies in the category. "Good social presence" goes. "841,971 Instagram followers" can stay if it matters to the decision.
- Read across the boxes for a pattern, as above. Four lists are not an analysis. One sentence that connects them is.
- Put the date of the next update in the diary. A SWOT without a second reading cannot show a change.
The grid is a sorting step, not a plan. Turning it into moves is the job of the TOWS matrix, which crosses the boxes and asks what to do about each pair.
What it cannot tell you
It cannot tell you why. The grid shows that fit and comfort lead the complaints. It does not show whether that is the shape of the product, a batch problem or buyers choosing the wrong size.
It cannot see what is not public. A new factory, a funding round that has not been announced, a rival about to launch: none of that is in the data. Threats like a new entrant or a regulation need other sources, and a grid that looks complete can hide that gap.
It cannot weigh the lines against each other. 79% of AI answers and a 2.58 on Trustpilot are both true. Which matters more depends on where your buyers decide, and the grid does not know that.
And it is a snapshot. A rating that moves by a few hundredths is noise. A line only counts as changed when the move is bigger than that measure's usual swing, which you only know after several readings.
FAQ
How do you do a SWOT analysis of a competitor?
Make the competitor the subject of all four boxes. Strengths and weaknesses come from their own numbers, opportunities and threats from the market and platforms around them. Put a source and a date on every line.
Should a competitor SWOT include my own company?
Not in the grid. Keep one subject, or the boxes stop meaning the same thing. Your own position comes in the next step, when you set their weaknesses against your strengths in a TOWS matrix or a battlecard.
How often should a competitor SWOT be updated?
As often as the numbers behind it move. Prices and ads can change weekly. Ratings and complaint themes move over months. Quarterly is a sensible floor.
Is a SWOT enough for a competitive analysis?
No. It sorts what you already know. The step-by-step competitive analysis covers how to collect it, and frameworks like the strategy canvas show things a SWOT cannot, such as where a rival chooses not to invest.
Pick one competitor. See what we find.
Your first benchmark is free. One domain, measured against you, no card and no call.