// pillar 03 · signals · built with Actions

Signals we observed, not inferred.

Job changes, hiring swings, funding fields, post activity — detected by scheduling LinkedIn Actions on the entities you watch, then diffing live reads in your stack. No packaged signals feed: same Actions API, same credits, same callbacks.

Edges is not related to LinkedIn and is not an official LinkedIn product.

your pipeline · schedule Actions · difflive
  • Mia Larsson joined Ramp as VP Productdiff · linkedin-extract-people-experiences12 s
  • Marble raised a $42M Series Bdiff · linkedin-extract-company · last_funding_*1 m
  • Linear opened 14 senior eng roles in 7 daysdiff · salesnavigator-extract-employees-count3 m
  • David Park reacted to your post — 4× this weeklinkedin-extract-people-reaction-activity6 m
  • Priya R. promoted to Head of Growth · Verceldiff · linkedin-extract-people11 m
  • Henry Q. viewed your team's profiles · 3 viewers this hourlinkedin-extract-profile-viewers18 m

// why this matters

Job-change timing is the highest-ROI signal in B2B.

Champion departures, new-hire windows, and response speed — the business case before the mechanics. The long thesis lives in The Intent Economy; this page is how you ship it on Actions.

Pipeline from job changes

~$1.2M

101 opps · 5,000+ tracked moves · year one

Outreach × UserGems

Contract ROI

First-year ROI on the job-change program

Outreach × UserGems

Champion conversion

Past-champion job-change leads vs cold

UserGems research

When a champion leaves, the account is at risk and the new company is often a warm door — if you notice. Manual LinkedIn monitoring does not scale past a few dozen accounts; a weekly CRM enrichment job finds the move after the window has closed.

Schedule profile and experiences extracts on the watchlist you already own, diff the live reads in your stack, and fire Slack, CRM tasks, or agent outbound on the same API key. That is the product below — not a packaged intent feed.

Sources are third-party research, not Edges metrics. See Outreach × UserGems · UserGems signals research.

// 8 patterns · Actions

Detect what changed — with the Actions you already call.

Schedule an extract or search, store the last snapshot, emit when the next run differs. Each card names the Action slug from the library.

Job change

Schedule profile or experiences extracts on your watchlist. Diff titles and companies in your app when a run returns.

linkedin-extract-people-experiences

Funding fields

Company extracts include last funding type, amount, date and investors. Re-read on a cron and alert when those fields move.

linkedin-extract-company

Hiring / headcount

Pull employee counts or search employees on a cadence. Spikes and drops are deltas you compute between runs.

salesnavigator-extract-employees-count

Engagement

Post reactions, comments and profile viewers — Action outputs you can roll up into warmth scores yourself.

linkedin-extract-people-reaction-activity

Profile viewers

Who looked at a connected identity’s profile. Schedule the viewers Action and treat new rows as warm inbound.

linkedin-extract-profile-viewers

Keyword / content

Search posts for topics or competitors, then enrich authors with people extracts in the same credit pool.

linkedin-search-content

Role posted

Search jobs filtered by company, seniority or function. New listings on accounts you watch are timing signals.

linkedin-search-jobs

Company moves

Re-extract company pages for domain, rebrand and affiliate shifts. Cross-check against your CRM account map.

linkedin-extract-company

// Delivery

Two ways Action results land.

Async and schedule modes share the same Action logic. Pick callbacks or pull outputs — documented in the runs guide.

async · schedule

Callbacks

Each input posts to your webhook when the run finishes — status, input, custom_data, results. Use custom_data to correlate CRM IDs. Spec: docs.edges.run/v1/runs/callbacks.

POST https://yours.app/webhook
{
  "run": { "run_uid": "…", "status": "SUCCEEDED" },
  "input": { "linkedin_profile_url": "…" },
  "custom_data": { "crm_id": "opp_42" },
  "results": [ /* action output rows */ ]
}
pull

Poll run outputs

Skip webhooks if you prefer: fetch outputs for a run_uid when your worker is ready. Same Action results either way — see execution modes.

GET /v1/runs/{run_uid}/outputs
{
  "results": [ … ],
  "cursor": "…"
}

// Playbooks

Three patterns teams ship within a week.

Schedule → extract → diff → enrich → notify. Your code owns the delta; Edges runs the LinkedIn reads.

Job-change alerts to AE Slack

Schedule experiences extracts on closed-lost buyers. When company or title changes, your worker posts Slack and opens the AE task.

Funding → ICP routing

Cron company extracts on your TAM. When last_funding_* fields change and the account matches ICP, enqueue outbound research.

Warm-prospect surfacing

Pull reaction and comment activity on your posts, enrich authors, score by seniority — a daily warm list from Actions, not a feed product.

Hiring-burst account picker

Track employee counts on target accounts. When headcount jumps in a function you care about, kick off account research jobs.

Keyword monitor → research

Schedule content search for competitor names or pain keywords, then linkedin-extract-people on authors and upsert into CRM.

Churn radar

Watch customer accounts with headcount + senior-profile extracts. Drops and departures fan out to CSM before renewal.

// FAQ

Signals, answered.

Cadence, callbacks, and credits — built on Actions, not a separate signals SKU.

01

What counts as a LinkedIn signal?

A signal is a change you detect by re-reading LinkedIn with Actions — a new role on a profile, a headcount swing on a company page, post activity, profile viewers, a job posting, funding fields on a company extract. Edges does not ship a separate signals feed or monitor product. You schedule the extract/search actions you need, compare results in your stack, and act when something moved.

02

How fresh is the data—how quickly after a job change will we know?

Freshness is the cadence you choose. Schedule linkedin-extract-people or linkedin-extract-people-experiences on the profiles you care about (hourly, daily, etc.), store the last snapshot, and emit an event when experiences or titles change. The same pattern works for company headcount via salesnavigator-extract-employees-count or company funding fields via linkedin-extract-company. Tighter cadence costs more credits; you control the trade-off.

03

Do you support webhooks?

Yes — for Action runs. Async and schedule modes deliver each input’s results to your callback URL (see docs.edges.run/v1/runs/callbacks). That is how you get near-real-time delivery after a scheduled extract finishes. Prefer polling? Pull outputs with GET /runs/{run_uid}/outputs instead of registering a webhook. There is no separate /v1/signals events stream.

04

How do Edges signals compare to Crustdata's firmographic feeds?

Crustdata ships broad firmographic and headcount datasets built for ranking and lead scoring — a feed you subscribe to. On Edges you build the trigger yourself: schedule the Actions that read the LinkedIn surfaces you care about, diff in your warehouse or app, then fan out. If you need a vendor-operated intent feed, a firmographic product may fit. If you want live LinkedIn reads on your own watchlist plus the same API for enrichment and outreach, build the signal layer on Actions.

05

Can I monitor just my CRM—only trigger on my accounts?

Yes — pass only the LinkedIn URLs (or Sales Nav IDs) from your open opportunities, customers, or ICP as inputs to scheduled Actions. Your app owns the watchlist; Edges runs the extracts. Add and remove inputs as CRM rows change. That keeps credit spend and noise tied to accounts you already care about.

06

Can signals drive an AI agent's outbound?

Yes. When a scheduled extract shows a job change or engagement spike, your agent runtime gets the Action callback (or a polled output), composes context-aware messaging, and closes the loop with Edges engagement Actions — connect, message, InMail — under the same API key and identity rate limits.

07

What's the difference between this page and the intent-economy content?

This page is the how: schedule Actions, deliver via callbacks or run-output polling, detect deltas in your code. The /intent-economy series is the why — why reactive, event-driven GTM beats weekly batch enrichment. Read the essay for the thesis; read the Action docs and this page to ship.

08

How do you price signals—per event, per monitor, or per credit?

There is no separate signals SKU. You pay the same credit model as every other Action: each scheduled or live extract/search consumes credits when it runs. Event volume is a function of how often you schedule and how many entities you watch. See pricing for the calculator, or talk to sales for high-cadence watchlists.

Catch the moment, not the lagging list.

Schedule the Actions that watch your ICP, diff in your stack, and fire outbound on the same API key.