think-swot
SWOT
SWOT sorts a situation into four boxes: internal Strengths and Weaknesses, external Opportunities and Threats. It is famous and frequently requested, and its evidence is weak: the most-cited field study found that bare SWOT grids "did not constitute analysis." This skill runs SWOT honestly. It leads with that caveat, then forces the three things a bare grid omits and that make it worth running at all: each factor tagged as evidence or assumption, each list prioritized to a few load-bearing items, and a TOWS matching step that converts the four lists into named strategic options. The output is a disciplined SWOT plus a TOWS option set, not a long undifferentiated list.
Before you run this: the evidence caveat
SWOT is tier X (weak or contradictory evidence). The single most-cited study of SWOT in practice, Terry Hill and Roy Westbrook, "SWOT Analysis: It's Time for a Product Recall" (Long Range Planning 30(1), 1997, 46-52), reviewed SWOT use by 14 consulting firms across more than 20 of 50 UK manufacturing firms and found the grids collapsed into long, undifferentiated lists, averaging over 40 general factors, with no prioritization, no verification of any point, and no connection to the rest of the strategy work. Their verdict: the activity "did not constitute analysis."
So a four-box list is not a deliverable. If you run SWOT here, the value is not the boxes, it is the discipline this skill adds on top: tagging each factor as evidence or assumption, cutting each list to the few that matter, and the TOWS matching step (Weihrich 1982) that turns the lists into options. If what you actually need is to decompose a situation rigorously, think-issue-tree is the stronger move; for parallel external futures, think-scenario-planning. Do not report a bare grid as analysis, and do not invent quantified claims about SWOT's effectiveness (none are supported).
When to Use
- The user asks for a SWOT by name and you should run it honestly rather than refuse.
- A quick shared snapshot of internal and external factors is genuinely wanted, and you will add the missing discipline (prioritize, verify, match).
- As a front end to option generation, where the TOWS matching step is the real payload.