Custom scripts and web scrapers

A small program that does one boring thing and does it the same way every time: collects competitors’ prices, pulls listings, moves data from an export into a spreadsheet. It is often the cheapest way to remove a routine — cheaper than bringing in a whole system.

How the work goes

  1. 1 Source We look at the source: how the pages are built, whether there is an API, how often the markup changes and what the request limits are
  2. 2 Prototype We build a prototype on a small sample and show you the result — before writing everything else. If the data is not what you need, you find out on day one, not on day ten
  3. 3 Resilience We make it resilient: retries, pauses between requests, careful behaviour under the source’s load and predictable behaviour on errors
  4. 4 Export We connect the export and the schedule: the result arrives by itself in a spreadsheet, a database or as a short report in your messenger
  5. 5 Handover We hand over the source code with a manual: how to run it, where to change the parameters and what to do if the source moves

What people usually ask us to collect

Not a complete list — these are the most common requests.

  • Competitors’ prices and availability — for your own list of items, with a history of changes by date

  • Listings from classifieds and marketplaces: cards, parameters, publication date, seller contacts

  • Public registries and directories, when you need to check a counterparty or build a list by criteria

  • Your own data from a system that has no proper export — including old archives

  • One-off processing of a large file: clean it, merge duplicates, sort it into categories, find discrepancies

  • A bridge script between two services that cannot talk to each other directly

Tell us what needs to be done

We reply within one working day. You talk to the engineers who will build the project.