Buying Signals Monitor — Hiring, Funding, Tech & Exec Triggers

The deal you lose is usually the one you were a month late to. They raised in March and picked a vendor in April. The new CRO arrived in May and re-signed everything by July. They ripped out your competitor's tool in June and nobody told you. None of that is hidden — it is public, scattered across five different places, and nobody on your team has time to check five places for two hundred accounts every week.

That is what this does. Give it your target accounts. Get back every buying signal, per company, in one call — scored, prioritised, and with the one thing a monitor is actually for: what changed since last time.

---

The five signals

Signal	What it answers	Why it is a trigger
Hiring	which roles are open right now	headcount goes up before budget goes out
Tech change	what they just started — or stopped — using	a tool arriving or leaving is a decision in progress
Funding	did they just raise, how much, from whom	the single strongest buying window there is
Ads	are they advertising, ramping, or paused	paid spend tracks growth pressure
Leadership	who just took a senior seat, and from where	a new leader re-picks their predecessor's stack

Each one becomes a block on the row: present, value, momentum, summary and full details. Together they become one 0–100 signalScore, a priority of hot / warm / watch / cold, a sellTo list and a whyNow line you can paste into a sequence.

---

What a monitor gives you that a scraper does not

Run it once and you get a snapshot. Run it on a schedule and every run also answers what moved:

• newSignalsSinceLastRun — the signals that started firing since you last looked
• lostSignalsSinceLastRun — the ones that stopped
• signalScoreChange and trend — heating, cooling, steady, new
• isNewCompany, runsSeen, firstSeen — how long this account has been on your radar

Turn on "only what changed since my last run" and each run hands you a short list of accounts that actually moved. That list is the product. Wire it to a weekly schedule and stop reading dashboards.

---

Who it is for

• Outbound and SDR teams — a weekly "who got hot" list for the accounts you already care about, with the opening line written.
• AI / GTM agents — one tool call, a flat dataset, no login and no key. signalScore, priority, whyNow, sellTo, newSignalsSinceLastRun are all on the row.
• Account managers and CS — know the day your champion leaves or a competitor's tool appears.
• Founders and investors — watch a market, not a spreadsheet.

---

How to use it

Watch your accounts:

{
  "companies": ["gong.io", "ramp.com", "lemlist.com"],
  "signals": ["hiring", "tech", "funding", "ads", "leadership"],
  "daysBack": 30
}

Find the market first, then watch it:

{
  "keywords": ["revenue intelligence software"],
  "locations": ["United States"],
  "companySizes": ["51-200", "201-500"],
  "maxCompanies": 50
}

The weekly monitor — only what moved, only what is worth a call:

{
  "companies": ["gong.io", "ramp.com"],
  "onlyChanged": true,
  "minSignalScore": 30
}

---

Scoring, in plain terms

Money in the bank beats a new decision-maker, which beats a stack change, which beats headcount growth, which beats paid media — so the weights are funding 28, leadership 24, tech 20, hiring 16, ads 12. Each signal is scored on its own terms first (a $5M round last week outranks a $200M round last quarter), then the total is normalised over the signals you enabled, so a two-signal run still uses the whole 0–100 range and rows stay comparable inside a run. No model call, no API key, and the same input always gives the same score.

---

Built to keep running

• Every signal is independent. If one source cannot answer, that column is omitted for that company, named in signalsSkipped, and the rest of the row ships. A run does not fail because one signal did.
• Quiet accounts still come back — with present: false, which is a different and more useful fact than a missing row. They are never charged for.
• Deduplicated by domain, so the same company on your list twice is fetched once.
• Bounded spend. maxSpendUsd is a hard ceiling on what the lookups may charge, and the run summary reports what each source actually cost.

---

Notes

• Dataset-only. No UI, no dashboard, no login. Four ready-made views: signals, detail, changes, audit.
• One memory per watch. The watch ID is derived from what you are watching, so the same list keeps the same history automatically — or pin your own to keep it stable while you edit the list.
• Public data only. Nothing behind a login is ever touched.