Your market is a technology, not an industry
A technology recruiter's patch is badly described by every field a company database offers. The companies running the platform you recruit around are spread across retail, logistics, banking and software, at every size. An industry filter finds a fraction of them and a great deal else.
What they share is the stack. A company that runs the technology employs people who know it, loses some of them, and replaces them. That makes the list of companies running it the nearest thing a specialist desk has to a territory: a fixed set of accounts you intend to know, call and revisit, whether or not any of them is advertising this month.
This play builds that territory once, properly. It is deliberately not about timing. When you want to know which of these accounts to ring this week, act when a company adopts a technology ranks them by recent adoption and live adverts. Run this one first and the signal play has a list to work from.
What you get back
A CSV with one row per company: name, domain, country, headcount, the source or sources that detected the technology, the dates it was first and last seen where that was confirmed, and a tier with the reason. For tier A, the engineering leader with title, LinkedIn and a checked work email.
Above it: how many companies matched in total, how many you pulled, how many fitted your desk, and how many came back with a named leader. The first number is the size of your market, and it is worth having even if you stop there.
Variations worth knowing
Require two technologies. The company search can demand that a company runs both of two technologies. That turns a very long list for a common platform into a short one for your real niche.
Narrow by funding stage. The same search filters on funding stage, for desks that only work venture-backed companies.
Widen from your winners. Once some of these accounts convert, find more companies like your best clients extends the list from them.
Where this goes wrong
A loose technology name. A common word matches unrelated products, and every company returned is charged. The name is confirmed against each source's catalogue before the pull.
A detection that has lapsed. A company that dropped the technology can stay on a list. The last-seen date is the check, which is why tier A is confirmed on a second source.
Sites are not companies. A list built from websites can count one group several times. Rows are merged on domain, and brands that look like one group are flagged for you to collapse.
Pulling the whole market. A popular technology matches more companies than a desk can work. Cap the pull at what you will call this year, and leave the rest counted but unbought.
