CLASSIFY: business-strategy — the task calls for structured strategic reasoning (market/competitive analysis), so Specificity-and-Context and Constraints-and-Format carry the most weight, with Structure-and-Reasoning close behind and Failure-Mode Prediction used to justify the fixes. STRUCTURE AND REASONING The prompt gives no analytical path: it doesn't say whether the model should ground each SWOT quadrant in market data, competitor comparison, or founder-stated assumptions, and it doesn't ask the model to justify why an item belongs in one quadrant over another (a weak team, for instance, could be framed as either a weakness or a threat). Without this, the model defaults to generic quadrant-listing rather than reasoned analysis. SPECIFICITY AND CONTEXT Several load-bearing terms are undefined. "AI-powered" doesn't say what the AI does (scheduling, summarization, workflow automation). "Remote team productivity" doesn't specify segment (SMB vs enterprise) or use case (async communication, task tracking, meeting overhead). "New" doesn't establish stage (pre-launch idea, MVP, funded, revenue-generating), which changes what a realistic strength or weakness even looks like. "U.S. market" is stated but nothing about competitive landscape, pricing model, or team size is given, so the model has to invent a company profile from nothing. CONSTRAINTS AND FORMAT No output length, no instruction on how many points per quadrant, no instruction on whether to include a synthesis or recommended strategic move after the four quadrants, and no exclusion of filler like generic "strong team" or "market is growing" statements that add no analytical value. FAILURE-MODE PREDICTION A weak model will produce a generic, textbook SWOT with four bullet quadrants of vague, interchangeable claims (e.g. "innovative technology" as a strength, "competition" as a threat) that could apply to almost any SaaS startup, with no grounding in a specific product, segment, or stage. CONVERGENCE Ranked gaps: 1, product and segment undefined — ambiguous terms resolved: no to yes ("AI-powered" and "remote team productivity" now given a concrete feature set and target segment). 2, startup stage undefined — context added: no to yes (pre-launch/MVP stage specified so strengths and weaknesses are realistic). 3, no reasoning path — reasoning structure specified: no to yes (each item must be justified against a stated assumption or market fact). 4, output format unconstrained — format specified: no to yes (fixed item count per quadrant, plus a closing synthesis line). 5, no success criterion — success criteria defined: no to yes (specificity and non-genericness named explicitly as the bar). ENHANCED PROMPT Act as a startup strategy consultant advising a pre-launch founder. The company is an AI-powered SaaS product for remote team productivity, where "AI-powered" means it uses AI to automate async status updates and summarize meeting notes, and the target segment is small-to-mid-size remote-first companies (20 to 200 employees) in the U.S. market. The founder has a working MVP and no paying customers yet. Perform a SWOT analysis. For each of the four quadrants (Strengths, Weaknesses, Opportunities, Threats) give exactly three items, and for each item write one sentence stating the point and one sentence justifying it against a specific market fact, competitor behavior, or stated fact about the company above — do not include a claim you cannot justify this way. After the four quadrants, add a short synthesis of two to three sentences naming the single most important strategic move this SWOT implies. Avoid generic claims that could apply to any SaaS startup (e.g. "strong team," "growing market") unless tied to the specific product and segment described above. Do not use bulleted or markdown-list formatting anywhere; write each quadrant as short labeled lines in flowing prose, and if a hyphen is used as a separator keep it tight against the text with no following space. The analysis succeeds if a reader unfamiliar with the prompt could not guess it was written from a generic template — every point should be traceable to the product, segment, or stage defined here. WHAT CHANGED The product and segment were defined in concrete terms, closing the failure mode where the model invents a vague, generic AI SaaS profile with no real market footing. A startup stage (pre-launch MVP, no customers) was added so the strengths and weaknesses reflect a realistic company rather than an idealized one. A justification requirement was attached to every SWOT item, forcing each point to trace back to a stated fact rather than floating as an unsupported generic claim. A fixed item count and a closing synthesis line were specified so the output has a predictable shape instead of an open-ended list of arbitrary length. A success criterion and an explicit ban on generic filler phrases were added so genericness itself becomes checkable rather than left to the model's discretion.