SWOT has a bad reputation among people who have sat through a lot of them, and the reputation is earned. The typical output is four boxes containing adjectives — “strong brand”, “limited resources”, “growing market”, “new entrants” — that nobody disagrees with and nobody acts on.

The framework is not the problem. The problem is that it is easy to fill in badly, and a badly filled SWOT looks identical to a good one at a glance.

What the four quadrants actually mean

Start with a distinction that is routinely lost: strengths and weaknesses are internal; opportunities and threats are external.

“The market is growing” is not a strength, it is an opportunity — it is true whether or not you exist. “We have a strong engineering team” is not an opportunity, it is a strength. This sounds pedantic and is not: mixing the axes is what produces boxes that all say roughly the same thing, and it destroys the framework's only real output, which is the interaction between quadrants.

Rule 1: Every entry needs a comparator

A strength is only a strength relative to someone. “Fast onboarding” means nothing on its own; “onboarding takes two days against a category norm of three weeks” is a finding.

Applying this rule kills roughly half of a typical SWOT immediately, which is the point. What survives is the material that could actually inform a decision.

The same discipline applies in reverse, and it is where most competitor SWOTs go wrong: a smaller team, less funding or a shorter history is not automatically a weakness. That inference fails the inversion test more often than it holds.

Rule 2: Every entry needs evidence

Attach a source to each line. Not a footnote for its own sake — a source is a forcing function. It is remarkably hard to write “weak customer support” once you have to say where you learned it.

Sources also let a reader weigh entries against each other. “Weak support, per 40 reviews mentioning response time across two review sites” and “weak support, per one conversation at a conference” are different claims that look identical without attribution.

Where the evidence comes from is covered in the public record.

Rule 3: The output is the interactions, not the boxes

This is the rule that turns SWOT from a describing tool into a deciding tool, and it is the one almost always skipped.

Once the four quadrants are filled, the analysis is in the pairings:

  • Strength + Opportunity. Where do we have an advantage that this external shift makes more valuable? These are your offensive moves.
  • Strength + Threat. What can we use to defend? These are your holding actions.
  • Weakness + Opportunity. What are we unable to capitalise on? These are the highest-value fixes, because there is a specific payoff attached.
  • Weakness + Threat. Where are we exposed? These are the risks that deserve a named owner.

A SWOT that stops before this step has produced an inventory. The inventory was never the deliverable.

Running one on a competitor rather than yourself

A competitor SWOT inverts the perspective and gains a constraint: you cannot see inside. Everything must be inferred from external evidence, which changes what each quadrant can contain.

  • Strengths come from what they demonstrably do well — consistent review praise on a specific dimension, technical capabilities visible in the product, an unusually strong hire.
  • Weaknesses come from repeated complaints, capability gaps, hiring that suggests a team stretched thin, or a support model that cannot scale with their pricing.
  • Opportunities and threats are the same external environment you face, read from their position instead of yours.

The discipline that matters here is separating observation from inference and saying which is which. “They posted eight support roles” is an observation. “They are struggling with support load” is an inference. Both are legitimate; conflating them is not.

The failure modes, named

  • Adjective soup. Entries that are opinions with no comparator and no source.
  • Axis confusion. External facts filed as internal ones.
  • Symmetry bias. Forcing four items into each box because it looks tidy. Real analyses are lopsided.
  • Politeness. Weaknesses softened until they are unactionable, usually when the SWOT is about your own company and the person who owns the weakness is in the room.
  • Stopping at the boxes. Covered above, and the most common of all.

When to reach for something else

SWOT is a general-purpose orienting tool. It is poor at questions of relative position across a market — for that, a positioning map shows in one image what four lists cannot. It is also poor at prioritisation, since it produces no ordering. Pair it with something that does.

Our Competitive Intelligence report runs SWOT this way for each competitor analysed — comparators stated, sources cited, inferences labelled, and the quadrant interactions carried through into recommendations rather than left as four lists.