Data extraction prompt for Founding Operators
The data extraction prompt adapted for founding operators. Body, role-specific tweaks, common pitfalls, and how to run it with Strawberry.

This is the data extraction prompt adapted for founding operators. It exists because spending too much time on admin, and the prompt below is the shape that actually survives contact with how founding operators work day to day.
What this prompt is for
Purpose: extract a structured dataset from messy webpages or PDFs - directory pages, public profiles, search results, scraped sites. For founding operators specifically, the value is that it turns a recurring admin task into a 5-minute repeatable artifact. This isn't a generic template - the items below are tuned for founding operators and the tools they actually live in.
The data extraction prompt (prompt)
- The source - a single URL, a list of URLs, or a directory pattern
- The schema you want back (CSV columns, JSON fields)
- The pagination strategy (next button, page number, infinite scroll)
- The validation rules (email format, phone format, dedup key)
- The limits (max rows, time budget, source pages to verify)
Adjustments for founding operators
founding operators typically live in . That changes how this prompt runs:
- Pull the inputs from the apps founding operators actually use, not generic SaaS exports.
- Anchor on recent activity in the prospect or company - it's the highest-signal field for this role.
- Skip items that don't apply to your weekly cadence; this is a starting shape, not a contract.
The most common way to mess this up
Skipping the schema - if you don't specify the shape, you get prose, not data. For founding operators, this shows up as spending the saved time on more admin instead of higher-leverage work. Build the prompt into your week, not as a one-off.
How Strawberry runs this prompt
Strawberry handles login walls, captchas, and pagination patterns that no-code tools can't - because it runs in your real browser. For founding operators, Strawberry uses your live tabs and connected apps - so the prompt is filled with your real context, not a placeholder.
When to use this, when to skip
Use this prompt when the work recurs (weekly, per-prospect, per-meeting). Skip it when the situation is novel and judgment-heavy - the prompt is a baseline, not a substitute for thinking.
Caveats
Strawberry holds back on sending email, updating CRM records, or changing shared systems until a human approves the action. Treat the agent as a fast first-draft author, not an autopilot.
Data extraction prompt
Step 1
the source - a single URL, a list of URLs, or a directory pattern
Step 2
the schema you want back (CSV columns, JSON fields)
Step 3
the pagination strategy (next button, page number, infinite scroll)
Step 4
the validation rules (email format, phone format, dedup key)
Step 5
the limits (max rows, time budget, source pages to verify)
FAQ
How long does this prompt take to fill out?
For founding operators, a first pass runs in 10-20 minutes. With Strawberry doing the data pulls, it drops to 2-5 minutes per artifact.
Can I customise this for my team?
Yes - the shape above is a starting point. Strip items that don't apply, add items that match your weekly cadence.
What is the biggest mistake?
Skipping the schema - if you don't specify the shape, you get prose, not data.