You have no reason to trust a team you found on the internet. So don't — judge the output instead. Tell us a public page and the fields you want, and we'll extract a hundred real rows from it and send you the sheet. Free, once per company.
Every scraping quote is a promise about data you haven't seen. A sample turns that promise into a spreadsheet you can open, sort and check against the live site yourself.
A URL and the columns you want. That's the whole brief.
Real data from the live site — normalised, deduplicated, exactly as the full job would arrive.
Within 24 hours, with a fixed quote for the full job alongside it.
Keep the sample either way. There is no follow-up sequence and no invoice.
One free sample per company. 100 rows, one source. Publicly accessible pages only — nothing behind a login we aren't authorised for, and nothing a site's terms forbid. If your source falls outside that, we'll say so within the hour rather than go quiet.
There isn't one, but here is the honest reasoning: most people who see accurate data from their own target site go on to order the full job, and the ones who don't cost us an hour. It's cheaper than advertising and more convincing than a case study.
No. You get the sample, the quote, and one follow-up a few days later. If you don't reply to that, we stop.
Not free. A hundred rows is enough to judge field accuracy, formatting and completeness — which is the entire point. Beyond that it becomes the job itself.
Then the sample is more useful, not less. If a source has aggressive bot protection or a structure that makes extraction unreliable, we'd rather find that out on our own time and tell you honestly than discover it after you've paid.
The sample is not a mock-up or a screenshot. It is working code run against your actual target, returning real data in the format you asked for.
You receive a working extraction covering 25 to 100 rows from your target source — enough to verify field accuracy, data completeness, and format before committing to the full run. Where relevant you also receive a short note on any technical challenges we identified (pagination, JavaScript rendering, rate limits) and how we handled them.
Most samples are delivered within one working day of receiving the brief. Complex targets with authentication or heavy JavaScript rendering may take up to two days. We tell you upfront if a target is unusually difficult.
It is free. There is no obligation to continue. We do it because it eliminates the single biggest risk in data extraction work — the risk that the data you expected does not exist in the form you assumed. A sample resolves that before any money changes hands.
If the sample looks right, we produce a fixed-price quote for the full run. If it does not look right, we discuss what needs to change. If the target turns out to be technically infeasible we tell you that and the engagement ends there — no charge.
Yes. For automation or AI agent work, the sample is a working proof-of-concept: a demonstrable workflow or a connected agent endpoint responding to a test prompt. Same principle — real output before commitment.