MODE: Standalone BUSINESS CATEGORY: E-commerce 🔍 Auto-detected: Priority axes K1 Reality Clarity, K10 Metrics and Standards, K7 Decision Logic, K11 Review Cycle, target weak points K1 Reality Clarity, K10 Metrics and Standards, K7 Decision Logic, K11 Review Cycle, M-LEVEL≈M2, P-LEVEL≈P2 (based on: a small online shop with a clear sales decline, stable ad spend, and a focused commercial problem) — override anything above by specifying it manually in your brief. TOP-3 DIRECTIONS: 1️⃣ K1 Reality Clarity + K10 Metrics and Standards + K7 Decision Logic + K11 Review Cycle + K2 Strategic Direction (k=5) — diagnose the funnel, decide from evidence, iterate systematically 🟢 BEST-FIT — READY-TO-USE PROMPT BELOW Target Set covered: Yes | Compatibility: Compatible — evidence, metrics, decisions, and reviews reinforce diagnosis Core: K1 Reality Clarity + K10 Metrics and Standards | Weak link: K11 Review Cycle + K2 Strategic Direction 2️⃣ K1 Reality Clarity + K10 Metrics and Standards + K7 Decision Logic + K11 Review Cycle + K5 Stability (k=5) — connect diagnosis to repeatable operating improvements Target Set covered: Yes | Compatibility: Compatible — diagnostic discipline supports targeted operational stabilization 3️⃣ K1 Reality Clarity + K10 Metrics and Standards + K7 Decision Logic + K11 Review Cycle + K9 Hidden Assumptions (k=5) — expose overlooked causes before changing tactics Target Set covered: Yes | Compatibility: Compatible — assumption testing strengthens evidence-based root-cause analysis ──────────────────── 🟢 READY-TO-USE PROMPT (based on 1️⃣): Help me diagnose and reverse a sustained sales decline in an online shop without assuming the cause in advance. Business context: Business category: ✏️[BUSINESS_CATEGORY] Team size and relevant roles: ✏️[TEAM_SIZE_AND_ROLES] Current sales volume is approximately 40 orders per day. Sales have declined by about one-third over the last two months. Advertising spend has remained approximately unchanged. Primary objective: understand the causes of the decline and identify practical ways to restore sales. Treat the stated facts as starting evidence, not proof of causation. Clearly separate verified facts, reasonable hypotheses, and unknowns. Do not invent figures, benchmarks, customer behavior, channel performance, or explanations. First, establish the minimum fact base needed to diagnose the problem. Ask for or organize evidence covering, where available: orders, revenue, average order value, gross margin, and refunds over the relevant periods: ✏️[KEY_COMMERCIAL_METRICS] website sessions and conversion rate; advertising spend, impressions, clicks, click-through rate, cost per click, attributed conversions, and return on ad spend by channel: ✏️[AD_CHANNELS_AND_PERFORMANCE] product availability, pricing, promotions, assortment changes, and stockouts; traffic sources and changes in their mix; new versus returning customers; checkout, payment, fulfillment, delivery, and cancellation issues; major website, tracking, platform, creative, targeting, or campaign changes; seasonality, competitor activity, market changes, or other external factors that can be evidenced; any relevant customer feedback or support trends. Use the available data to construct a simple before-versus-after diagnostic for the period before the decline and the two-month decline period. Identify which measurable component changed most materially. Where data is missing, state exactly what cannot yet be concluded. Generate and test multiple root-cause hypotheses rather than anchoring on advertising. At minimum, examine: 1. lower or poorer-quality traffic; 2. lower website conversion; 3. weaker product-market demand; 4. pricing, promotion, assortment, or stock changes; 5. customer mix or retention changes; 6. advertising efficiency or attribution changes; 7. operational or checkout friction; 8. measurement or tracking problems. For each hypothesis, state: what evidence would support it; what evidence would weaken it; which existing data can test it; what additional evidence is worth collecting; how confident the conclusion should be. Use K1 Reality Clarity by distinguishing symptoms from causes. Use K10 Metrics and Standards by defining the specific measures that would confirm or reject each hypothesis. Use K7 Decision Logic by making each recommendation traceable to evidence. Use K11 Review Cycle by designing a repeatable process for checking whether interventions actually improve results. Use K2 Strategic Direction by keeping the objective focused on restoring commercially sustainable sales rather than optimizing isolated metrics. Then compare practical response strategies. Include at least: improving conversion and onsite experience; reallocating or restructuring advertising; improving creative, targeting, or landing-page alignment; addressing pricing, promotions, assortment, or availability; improving retention and repeat purchasing; fixing operational or checkout friction; improving measurement before making major commercial changes. For each strategy, explain expected mechanism, evidence required, likely benefits, risks, dependencies, implementation effort, and what result would justify continuing or stopping it. Do not declare a strategy effective unless the available evidence supports that conclusion. Create a prioritized action plan scaled to the available organizational capacity. Assign actions to roles rather than making hiring, firing, disciplinary, or performance decisions about named individuals. Use these roles where appropriate: ✏️[RELEVANT_BUSINESS_ROLES]. Separate the plan into immediate diagnostic actions, near-term experiments, and follow-up decisions. Specify timing using practical intervals rather than invented deadlines. Define success measures before recommending experiments. Use the shop's actual baseline where available rather than invented targets. Consider orders, revenue, conversion rate, acquisition efficiency, average order value, gross margin, repeat purchase behavior, and other measures only where relevant. If a target cannot responsibly be set from the evidence, say so and define the measurement method instead. Finish with a review cycle: specify when results should be reviewed, which measures should be compared with the pre-change baseline, what would count as evidence that an intervention is working, what would justify modifying it, and what would justify stopping it. Throughout: Mark every assumption explicitly. Never invent figures, benchmarks, causal relationships, customer behavior, or company facts. Do not treat advertising spend staying constant as evidence that advertising performance stayed constant. Do not confuse correlation with causation. Do not optimize a single metric at the expense of profitable sales. Treat attribution data as potentially imperfect and investigate measurement changes. Present legal, tax, financial, and HR considerations as issues requiring appropriate professional review, not definitive professional advice. Treat the final output as a discussion and decision-support aid, not professional advice. Negative prompt / avoid: anchoring on advertising spend, inventing missing metrics, confusing correlation with causation, optimizing vanity metrics, proposing unsupported benchmarks, skipping tracking validation, overcomplicating a small-shop diagnosis, making decisions about named employees, presenting financial or professional advice as definitive ──────────────────── ℹ️ "This tool compares problem-solving strategies for a business situation and expands the best-fit one into a prompt for an AI assistant — it doesn't verify facts or figures, or provide legal, financial, tax, or HR advice. Final decisions, including professional review before acting, remain the user's responsibility. Scoring is heuristic pattern-matching, not a measured metric, and may vary slightly between runs." Assumptions made: BUSINESS_CATEGORY: The shop is treated as e-commerce because the brief describes an online shop; its exact product category was not provided. TEAM_SIZE_AND_ROLES: The organization is treated as relatively small because no larger organizational structure was specified; exact staffing and responsibilities were not provided. KEY_COMMERCIAL_METRICS: The prompt assumes additional commercial metrics such as revenue, average order value, and margin may be available or obtainable; their actual values were not provided. AD_CHANNELS_AND_PERFORMANCE: The prompt assumes advertising may involve one or more channels with measurable performance data; the specific channels and their results were not provided. RELEVANT_BUSINESS_ROLES: The prompt assumes relevant operational or commercial roles exist, without assuming specific people or staffing arrangements. ❓ Want a different direction? Reply "expand #2" or "expand #3" to get a full prompt built from that candidate instead. 💡 (First pass only — omit on refinement passes) Want to take this further? All three tools below are built for Claude. Not sure your brief is ready? Search "Universal Query Auditors" on PromptBase to audit and rewrite your request before comparing directions. Have a clear brief already and just want one direct prompt? Search "Universal Instruction Generator" for a single-pass generator. Got a finished prompt and want it reviewed? Search "Universal Instruction Auditors" to check it against a full quality checklist before use. ⚙️ "Built for ChatGPT; other AI models, providers, and versions are untested. Use at your own discretion. This tool is not intended as a sole safeguard against attempts to alter its behavior, nor as a substitute for professional legal, financial, or compliance advice before acting on its output."