Data and analytics AI prompts are precise instructions that get Claude or GPT to reason about numbers without hand-waving. On Prompt Dock they cover real tasks: writing and explaining SQL, interpreting metrics, building analysis plans, cleaning and structuring messy data descriptions, generating dashboard summaries, and translating a business question into the right query. Every prompt is human-reviewed before it goes live, so the analytical steps are explicit rather than fuzzy. Free and premium prompts are both available, and you can reveal any prompt and run it live in the Playground on your own schema or figures so the output holds up before it informs a decision.
8 prompts
Describe a user interaction you want to track. Get a complete, engineering-ready tracking plan: canonical event name, exact trigger, a property table, what NOT to track, and a QA checklist that includes the error case.
Describe what you need to know in plain English and paste your schema. Get a correct, readable SQL query with CTEs, aliases, a plain-English explanation, and an index note — no guessing at column names.
Paste your slow query and its EXPLAIN ANALYZE plan. Get a root-cause diagnosis, exact CREATE INDEX statements, a query rewrite if the structure is the problem, and a clear 'what NOT to do' warning.
Paste a sample of your customer or user table. Get a set of named, mutually-useful cohorts with exact filter logic, the SQL or rule for each, the one action each cohort warrants, and a note on overlap — built for retention, expansion, or win-back, not vanity slicing.
Paste raw numbers from your analytics tool or a spreadsheet export. Get a tight executive summary: the three findings that matter, what each means for the business, and one assignable action per finding — under 400 words.
Describe a table and what it feeds. Get a prioritized list of data quality checks — uniqueness, nulls, ranges, referential integrity, freshness, distribution drift — as runnable SQL assertions plus dbt-test / Great Expectations equivalents, so bad data fails loud before it hits a dashboard.
Paste your control vs. treatment results — metrics, sample sizes, duration. Get a structured readout with statistical and practical-significance assessment, a ship/hold recommendation, and a caveats section that hunts for sample ratio mismatch and other traps.
Paste the header and first 20 rows of a messy CSV. Get a column-by-column quality profile, an ordered cleaning plan, and a complete, runnable pandas script with before/after logging and TODO stubs where a human decision is required.
Prompt Dock's Data & Analytics category gathers them, with every prompt human-reviewed before going live. Reveal the instructions, choose free or premium, and run each in the Playground on your schema or metrics, so you can verify the SQL or interpretation on real data before trusting it in a report.
Yes. Several data prompts are built to translate plain-English questions into SQL and to explain existing queries. Each is human-reviewed for correctness of approach, and you can paste your table structure into the Playground and run the prompt to see the query on your actual schema before using it.
Claude is strong for explaining logic and reasoning through multi-step analysis, while GPT is quick for generating queries and summaries. Data & Analytics prompts note a recommended model, and running one in the Playground lets you compare how each handles your specific data question before you decide.