The problem this solves
Job postings are the most underused public document a company publishes. Every role a competitor opens names the tools the hire will use, the initiatives they will drive, and the capabilities the company lacks today. A cluster of postings mentioning a new analytics stack, a first hire for a new market, a sudden run of partnerships roles - each is a spending decision announced in plain text. Reading them systematically across a rival's whole careers page is the part nobody has time for.
Intent signals compound the picture: the topics a company's people are actively researching say what problems it is trying to solve right now. Held together, hiring and intent data sketch next quarter's roadmap months before any announcement. Held separately, or read casually, they are just noise. The gap between teams that mine this and teams that skim it is a quarter of lead time on every competitive response.
How the mission runs
- Collect the hiring record. Hiring Data pulls the competitor's open roles and recent postings: titles, departments, locations, and full descriptions. Volume and mix get baselined - which functions are growing and which are frozen - because the pattern across postings carries more signal than any single role.
- Mine the descriptions for tells. The agent reads every description for named tools, platforms, methodologies, and initiative language. Requirements sections are effectively a tech-stack confession, and responsibilities sections describe projects that do not exist yet. Each tell is logged with the exact posting it came from.
- Layer in intent signals. Intent Data adds the topics the company is actively researching - categories it is consuming content about at above-baseline levels. Where intent topics and hiring tells point at the same initiative, the agent marks the inference high-confidence; where they diverge, it says so plainly.
- Write the forward read. The signals get synthesized into a forward-looking analysis: the initiatives the competitor appears to be funding, the tools they are likely adopting or replacing, and what each move would mean for your positioning if it lands. Every inference cites its postings and topics.
- Log it to Notion. The full analysis is written to a Notion page: signal tables with sources, the initiative-by-initiative read, and a confidence rating on each conclusion. Re-running the mission updates the page and highlights new postings, closed roles, and intent shifts since the last pass.
The prompt
This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.
What comes back
A structured Notion page holding the complete analysis: a hiring-signal table with quotes from the postings, the intent-topic read, an initiative-level forecast of what the competitor is investing in next quarter, and a confidence rating with cited evidence for every inference. It is written to brief an executive in five minutes, with the source trail intact for anyone who wants to pressure-test a conclusion.
Make it yours
- Watch three competitors at once and ask for a shared initiatives board, so parallel bets across the market - everyone hiring for the same capability - stand out immediately.
- Focus the mission on a single department, such as engineering or sales, when you care about one dimension of the competitor's build-out.
- Schedule it monthly and have the page open with a what-changed section: new roles, filled roles, and intent topics that spiked since the previous run.
Frequently asked questions
How reliable are inferences drawn from job posts?
They are inferences and the page treats them that way. A named tool in a requirements list is strong evidence; a projected initiative from one posting is weaker, and the confidence rating reflects that. Every conclusion links its source postings so you can judge the leap yourself.
What exactly does the intent layer add?
Intent Data reports the topics a company is researching at unusual volume, which corroborates or challenges the hiring read. A rival hiring data engineers while spiking on warehouse-platform topics is a much firmer signal than either fact alone, and the analysis calls out those intersections.
How often should this run?
Monthly fits most categories. Job postings turn over on roughly that cycle, and intent topics need a few weeks to show a real trend versus a blip. The Notion page is designed for repeated runs, accumulating a timeline of the competitor's build-out rather than a one-off snapshot.