Scenario-Based Data Analysis: Why Pre-Built Analysis Scenarios Beat Free-Form Exploration
What is scenario-based data analysis? This guide covers DataFish's 27 pre-built scenarios, compares scenario-driven vs. free-form analysis, and shows why 'scenario first, chat second' is the most efficient workflow.
What's Wrong with Free-Form Analysis?
Many people think the ultimate form of data analysis is 'ask anything, get answers' — upload data, ask whatever you want, AI responds. Sounds ideal, but there's a key problem in practice: you don't know what to ask. This is 'analysis blank page syndrome' — facing a dataset without knowing where to start. Symptoms: open a data tool, stare at data, not knowing the first question. Ask a few questions but feel directionless, jumping between topics. Finish analysis but unsure if you missed something important. Root cause: most non-data professionals lack systematic analysis frameworks — they know they should 'look at data' but not how, what to look at, or in what order.
How Does Scenario-Based Analysis Solve This?
Core idea: package data analysis experts' experience into pre-built scenarios — you don't need to know how to analyze, just tell AI what problem you want to solve. Example: You run a tea chain and want to know 'which products are most profitable.' Free-form: ask 'analyze my products' and get a generic response. Scenario-based: select 'Product Contribution Matrix,' and AI automatically runs a complete Boston Matrix analysis — calculating each product's volume, growth rate, margin, auto-classifying into Star/Cash Cow/Question Mark/Dog quadrants, with optimization recommendations. The difference: free-form is 'open-ended questioning,' scenario-based is 'goal-oriented analysis framework.'
What Do DataFish's 27 Pre-Built Scenarios Cover?
By industry: F&B (5) — channel profit analysis, product contribution matrix, waste black hole, traffic bottleneck, customer retention curve. Beverage/Coffee (5) — hit product dependency, category margin, hourly revenue/sqm, member repurchase frequency, combo mining. Retail (5) — SKU velocity, ABC inventory, basket analysis, shelf efficiency, member tier analysis. Fresh Food (4) — weather-linked procurement, full-chain waste tracking, freshness pricing, seasonal windows. Fashion (4) — bestseller gene analysis, size distribution, lifecycle management, try-on conversion. General (4) — business diagnosis, weekly/monthly report, anomaly detection, target attainment. Each scenario includes: what data is needed, what dimensions to analyze, what results to output, what recommendations to give.
'Scenario First, Chat Second' Workflow: The Most Efficient Method
Best practice isn't 'only scenarios' or 'only free-form' — it's combining both. Workflow: Step 1, select a scenario — after upload, choose the best-matching analysis scenario. E.g., 'Business Diagnosis' for overall store health. Step 2, review results — AI completes analysis in 3-5 minutes with a full report (charts + insights + recommendations). Step 3, follow up — use chat to drill into details. 'Why does Store A have the lowest revenue/sqm?' 'Excluding promotions, what's the real growth?' 'Show me Store B's customer retention trend over 3 months.' Step 4, export — export the complete analysis as a report. Advantage: scenario analysis builds holistic understanding. Chat drills into specifics. The entire process is directed, focused, and comprehensive.
Case Comparison: Same Question, Two Methods, Different Results
Question: A chain owner wants to know 'how's business this month?' Method 1: Free-form. Owner asks AI 'how's business?' AI gives some basic metrics. Owner asks 'any anomalies?' AI lists some changes. But the owner isn't sure they asked the right questions or missed important things. 20 minutes, scattered results, no systematic coverage. Method 2: Scenario + chat. Owner selects 'Business Diagnosis,' AI runs 5-dimension full diagnosis (revenue, traffic, efficiency, products, customers) in 3 minutes. Report shows composite health score 78, Store C red-flagged (3 months declining traffic). Owner asks 'why is Store C traffic declining?' AI finds 2 competitors opened nearby, siphoning 30% of lunch traffic. Owner creates an action plan immediately. 8 minutes total, found a critical issue that might have been missed.
How to Start with Scenario-Based Analysis?
Step 1: Identify your industry and role. DataFish covers 6 industries (F&B, beverage, retail, fresh food, fashion, general). Step 2: Pick your first scenario. For first-time users, start with 'Business Diagnosis' — a full-dimensional health check for quick overview. Step 3: Prepare data. Different scenarios need slightly different data, but most just need basic sales data (date, store, category, amount, quantity). Step 4: Upload and analyze. Select scenario, upload data, AI auto-completes analysis. Step 5: Follow up with chat on details you care about. DataFish's new users typically go from 'I don't know how to analyze data' to 'I found a key business issue' in their first session.