The recruiter who publishes the numbers owns the conversation
Every niche has someone clients treat as the person who knows the market. Usually that person is simply the one who says, each month, what the market did: which titles were advertised, where, by whom, and what moved. The raw material is public. Employers publish it themselves, as job adverts.
What turns a job search into a report is discipline, not data. The same titles, locations, window and boards every month, so this month's figure can sit next to last month's. Then a plain statement of what the figures are: adverts counted on named boards inside a window, with agencies and duplicates removed. Not vacancies, not hires, not the whole market.
That honesty is what makes it quotable. A number with its source and date beside it survives a sceptical client. A percentage with no base does not.
What you get back
One page. Adverts by title, by location and by employer for the window. The employers advertising most in your discipline. A change column against the last report: employers that appeared, employers that went quiet, titles that moved. Market-wide totals from a job-data source, shown beside the board counts. Under the tables sit the definition, the caps, the sources, the pull date, and how many agency adverts and duplicates were removed.
A short write-up in your chosen format comes with it as a draft. Nothing is posted or sent.
Variations worth knowing
One client's view. Run the same definition across a single client's competitors and the report becomes an account review: who they are up against for the same people.
Pay, where adverts state it. Some adverts carry a salary. The report can list what was stated, next to the count of adverts that stated anything, which is often small.
From report to pipeline. The employers at the top of the table are leads. Find every company hiring for your specialism this week turns them into hiring managers, and pitch the hiring managers whose roles have been open too long works the adverts that have sat.
Where this goes wrong
A capped pull is a sample. If a pull reaches its cap, the split by title describes your pull, not the board. The report flags every pull that hit the cap.
Changing the definition. Add a title in March and March is up on February for no reason in the market.
Not every board takes a window. LinkedIn and Indeed searches accept a posting window. The German federal board and a company's careers page return what is live today, so those counts are a snapshot.
Agencies inflate the count. In some niches most adverts are other recruiters. Removing them by name and wording is good, not perfect.
Adverts are not demand. Roles filled through networks never appear, and one role can be advertised three times.
