SWOT analysis with a TOWS strategy matrix

Four lists decide nothing by themselves. The work starts when the internal is crossed with the external — the move Heinz Weihrich described in 1982 and called the TOWS matrix. There are four crossings, each with its own job.

SO: strength × opportunity

Your bets. “Rollout in two days” × “incumbents raised prices 40%” equals a migration campaign built on a promised timeline. Return on effort is usually highest here, and these are the pairs worth turning into a quarterly plan.

ST: strength × threat

The cheapest moves in the matrix, because you already hold what you need. “A competitor gave the feature away” × “support replies in eight minutes” equals moving the conversation off the feature and onto how fast the problem gets solved.

WO: weakness × opportunity

Projects live here, not tasks. “No Salesforce integration” × “budgets grew 20%” means the opportunity cannot be used until the integration exists. It is worth working out honestly whether you can fix the weakness before the window closes.

WT: weakness × threat

Risks. “Pipeline runs through the founder” × “paid traffic up 40%” is the scenario in which growth stops. With each of these you do one of two things: reduce it in advance, or say out loud that you are accepting it. There is no third option — or rather there is, “write it down and forget”, and it gets expensive later.

• Plans in opportunities. The most common one. “We could ship a mobile app” is not an opportunity, it is your idea; the opportunity is “half the traffic in this category is already mobile”. The test is simple: an opportunity exists even if you do nothing. The tool flags these by words like “we” and “our team”.
• The market in strengths. The reverse mistake: “a growing market” is not your strength, you did not build it. A strength is something you control and others lack.
• Items with no numbers. “Strong team”, “good product”, “high competition” are feelings. You cannot argue with them and you cannot check them against reality next quarter. One number turns an item into an argument.
• Stopping at four lists. The saddest one: the work is done and no decision came out of it. If a SWOT produced no action with a date attached, it did not need doing.

A fifth thing, not a mistake but a property: a SWOT almost always comes out flattering. People write more strengths and opportunities than weaknesses and threats, which is why the tool shows the plus-to-minus ratio separately. One item under weaknesses against five under strengths is not a description of the company — it is a description of the meeting where the table was filled in.

The method has serious and fair criticism, and knowing it is more useful than filling in the squares one more time.

• When you need to choose between options. A SWOT does not compare alternatives and does not size an effect. It describes a position, not a priority order. For choosing between tasks there is RICE and ICE; for allocating your own week, the Eisenhower matrix.
• When one person does it from memory. Then it is a list of beliefs. The minimum data: lost-deal reasons from the CRM, churn reasons, acquisition cost by channel, rollout times, competitors’ published numbers.
• When a box has more than seven items. Twenty-eight lines cannot be prioritised. That is no longer a decision tool, it is meeting minutes.
• When a “strength” has never been tested by the market. The classic criticism from Terry Hill and Roy Westbrook (Long Range Planning, 1997): across the companies they studied, SWOT produced long unsorted lists with no consequences for decisions. A strength customers will not pay for is just a characteristic.
• When the decision is already made. A SWOT is easy to bend towards a conclusion: add two opportunities, drop a threat. If the analysis exists to justify something, it is more honest to write the justification.
• When the subject is a product, not a company. For product decisions it is more useful to see what users treat as must-have and what merely delights them — that is the Kano model survey analyzer. A SWOT works at the level of the business and its market.

The practical conclusion: a SWOT is good as a warm-up before a decision, and as a way of saying out loud the things a team quietly disagrees about. It is bad as a conclusion. One hour to fill in, then the crossings and two or three decisions — otherwise the table stays in the deck.

“Show an example” fills all four boxes with an analysis of a small B2B SaaS company. It is deliberately uncomfortable: it contains the weaknesses nobody enjoys writing down, and numbers instead of adjectives.

The logic runs like this. Strengths hold retention above the category, the rollout time, proprietary data and support speed — all with numbers, all inside the company. Weaknesses hold the missing Salesforce integration, a pipeline dependent on the founder, acquisition cost against contract value, and missing certifications: each one cost specific deals. Opportunities hold nothing but external shifts — incumbents repricing, budgets growing, a new channel, changing buying habits. Threats hold a free feature from a large competitor, expensive traffic, customer concentration in one industry, and possible data-residency requirements.

From that set the matrix produces visible pairs immediately: “rollout in two days” against incumbents repricing is a campaign you can brief; “no Salesforce integration” against growing budgets is a project without which the money cannot be taken; “pipeline through the founder” against expensive traffic is a risk worth reducing before it fires.

1. Weihrich H. The TOWS Matrix — A Tool for Situational Analysis, Long Range Planning, 1982 — the SO / ST / WO / WT crossing.
2. Learned E., Christensen C., Andrews K., Guth W. Business Policy: Text and Cases, 1965 — the LCAG model the internal/external split grew out of.
3. Hill T., Westbrook R. SWOT Analysis: It’s Time for a Product Recall, Long Range Planning, 1997 — a study of 20 companies: long unstructured lists with no effect on decisions.
4. Panagiotou G. Bringing SWOT into Focus, Business Strategy Review, 2003 — why a SWOT without external data turns into a set of beliefs.
5. Humphrey A. SWOT Analysis for Management Consulting, SRI Alumni Newsletter, 2005 — a first-hand account of where the method came from at Stanford Research Institute.