Company Change-Event Feed - GTM Signal Delta Tracker

🔄 What can Company Change-Event Feed do?

Give it a company domain and it returns only what changed since your last run: typed change events across hiring, tech stack, funding, firmographics and social presence. One flat row per company.

It is an orchestrator. It does no scraping of its own. It calls fleet sub-actors, normalizes their output into one company state, stores a per-domain snapshot, and on the next run returns the deltas.

📦 What you get	⚙️ Features and integrations
🔄 Only the deltas, not a full re-dump 🚦 Severity and confidence on every event 🔍 old_value and new_value plus the source Actor ID 🧾 10 flat fields, one row per company	📡 Five GTM signal sources in one row 🛟 Partial-failure safe, a dead source never reads as a removal 💾 Snapshot returned to you, or stored per domain ⬇️ Export to JSON, CSV, Excel, HTML or XML

Bought by RevOps teams, outbound agencies and agent builders who need delta intelligence on target accounts rather than a fresh full profile every month.

🚫 The first run for a domain is a baseline and returns no changes. That is not a fault. It records the state to diff against. Every run after that returns the change events since the prior snapshot.

💡 Why use Company Change-Event Feed?

If you want	Read these fields
A trigger worth acting on today	has_high_severity, changes
Whether anything happened at all	total_changes, is_baseline
Recency	latest_change_date
Whether the run was complete	source_status
To hold the state yourself	snapshot

🛟 A missed fetch never looks like a removal

Each source is isolated. If one sub-actor errors or times out, the others still produce events, the failed source is marked degraded in source_status, and its prior state is carried forward.

This is the difference between a delta feed you can automate against and one you cannot. Without it, a sub-actor timing out would emit "removed their CRM" and fire an alert on a company that changed nothing.

Read source_status before you act on an absence.

▶️ Ready-to-run examples

Example	What it does
High-Severity Account Signal Trigger	Alert only on high-severity changes
Hiring Ramp Signal Monitor	Watch for hiring acceleration
Tech Stack Adoption Change Alert	Fire when a company adopts new tooling
Track Competitor Website Changes	Monitor competitors for changes
Domain Change Tracker	Track changes on a watched domain

📋 What data can Company Change-Event Feed extract?

10 fields per company, with the detail inside changes.

Field	What it holds
changes	The typed change events, each with old_value, new_value, severity, confidence and the immutable Actor ID of the source that produced it
total_changes	How many events this run produced
has_high_severity	The one-field alert gate
latest_change_date	Most recent change in this batch
is_baseline	true on the first run for a domain
source_status	Per source: ok, degraded, skipped or not_selected
snapshot	Current state, returned so you can store it yourself
company_domain, company_name	The identity
run_date	When the run happened

⚠️ Severity is there so you can ignore most of it. CRM adoption, a hiring ramp and a funding round are high severity. Follower drift and an HQ relocation are low. A feed where everything is an alert is a feed nobody reads, so filter on has_high_severity first and treat the rest as context.

🎛️ Choosing sources and a severity floor

Two settings shape a run, and both were fixed before 2026-08-22: every run paid for five sub actor runs, and every change came back regardless of how small it was.

{ "domain": "stripe.com", "sources": ["hiring", "funding"], "min_severity": "high" }

sources picks which of the five to monitor: hiring, tech_stack, firmographic, funding, social. Leave it empty for all five. A source you do not select is reported as not_selected in source_status, which is deliberately a different value from degraded. degraded means the actor tried and could not read it. not_selected means nobody asked, so the absence of a change there is not a claim that nothing changed.

min_severity sets the floor on the changes array:

Value	What survives
low	Everything. The historical behavior and still the default.
medium	Material changes: role count moves, headcount jumps, tool changes, funding, exec moves.
high	Major changes only: funding rounds, acquisitions, exec moves, a CRM or sequencer appearing or disappearing, a large hiring ramp.

high is the setting for an alerting feed. A change carrying a severity the actor does not recognize is kept rather than dropped, so a labelling bug can never silently hide an event.

Both settings come back on the row as sources_requested and min_severity, so a row with an empty changes array is never confused with a row where only major changes were asked for.

🛠️ How to track changes at a company

1. Open the Input tab and put a bare domain in domain.
2. Run it once. That run is your baseline and returns no changes.
3. Run it again on your chosen cadence, weekly or monthly.
4. Filter on has_high_severity and read changes.
5. Check source_status before you act on something disappearing.

🧪 Using it in Clay

Add an Apify enrichment column and map your domain column to domain, which is the per-row enrichment shape. domains[] exists for batch runs outside Clay.

Because the actor stores a per-domain snapshot itself, you do not have to persist anything. If you would rather hold the state, take snapshot from the output and pass it back as previous_snapshot on the next run.

⏱️ Choosing a cadence

Monthly suits firmographics and social. Weekly suits hiring and tech stack, which move faster. Running it more often than the underlying sources change produces empty runs you still pay for.

💵 How much does it cost to track a company's changes?

You are charged once per company analyzed by this actor, and separately by each sub-actor it calls. Both parts are on your account. The worked example below is the number that matters.

Plan	Per company	Per 1,000 companies
Free	$0.06	$60.00
Bronze	$0.057	$57.00
Silver	$0.054	$54.00
Gold	$0.051	$51.00

There is also an Actor start event at $0.00005, charged once per run per GB of memory.

💳 This is an orchestrator, so the sub-actors it calls bill you too. The price above is for this actor. Each of the five fleet sub-actors it invokes charges its own per-domain event against the token that started the run, which is yours. Budget for the orchestrator plus the sources, not the orchestrator alone. A baseline run costs the same as a delta run, because it does the same work.

🧾 What a run really costs, measured end to end

The five sub-actors cost more than this actor does. Here is a real run rather than an estimate. Measured 2026-08-12 on a single company.

Charged to your Apify account	Cost
This actor, 1 company plus the start event	$0.0600
gtm-hiring-signal-scraper, $0.05 per company	$0.0501
funding-press-signal-scanner, $0.03 per company	$0.0301
company-social-presence-mapper, $0.015 per company	$0.0150
gtm-tech-stack-signal-scraper, $0.015 per company	$0.0150
company-firmographic-enricher, $0.004 per domain	$0.0040
Total for that run	$0.1743

The five sub-actors were 66% of that bill. The effective price was $0.174 per company, not the $0.06 in the table above, so budget roughly three times the headline rate for a full run.

sources changes that ratio, and it is the single biggest lever on this bill. As of 2026-08-22 you choose which sources to monitor, and a source you do not select is not run and not charged. Watching hiring alone on the numbers above is $0.0600 plus $0.0501, about $0.11 instead of $0.174. Watching firmographics alone is $0.064.

The default is still all five, so an existing task, saved input or schedule costs exactly what it cost before.

⌨️ Input

Everything is on the Input tab. The options worth explaining:

Field	Type	Default	What it does
domain	string	none	Bare company domain. The per-row enrichment shape.
company_name	string	none	Improves attribution on the news and identity sources.
domains	array	none	Batch mode.
previous_snapshot	object	none	Diff against a snapshot you hold rather than the stored one. Single company only.
sources	array	all five	Which sources to monitor. Unselected sources are not run and not charged.
min_severity	string	low	Report only changes at or above this severity.
sub_actor_timeout_secs	integer	90	Per sub-actor ceiling. Lower it for a quick test.

📤 Output

One flat row per company, exportable as JSON, CSV, Excel, HTML or XML.

{
  "company_domain": "stripe.com",
  "company_name": "Stripe",
  "is_baseline": false,
  "total_changes": 3,
  "has_high_severity": true,
  "latest_change_date": "2026-06-21",
  "changes": [
    {
      "type": "tech_stack_added",
      "old_value": null,
      "new_value": "HubSpot",
      "severity": "high",
      "confidence": 0.9,
      "source_actor_id": "qyd7nNyqFPelQViBx"
    }
  ],
  "source_status": {
    "hiring": "ok",
    "tech_stack": "ok",
    "funding": "ok",
    "firmographics": "ok",
    "social": "degraded"
  },
  "run_date": "2026-06-21T16:20:11Z"
}

💡 Tips

• Alert on has_high_severity only. Store the rest as account context.
• Every event carries the Actor ID of the source that produced it, so you can always trace a change back to the actor that saw it and re-run just that one.
• Lower sub_actor_timeout_secs when you are testing. The default is tuned for completeness, not speed.
• Do not treat a degraded source as "nothing changed there". It means nobody looked.

⚠️ Known limits

Runs take longer than a single-source actor. This actor waits on up to five sub-actors running in parallel, so a typical run is about 60 to 180 seconds. That is expected for an orchestrator and does not indicate a fault.

Free-plan sub-actors can hit their own caps. If you are on the free Apify plan, a sub-actor may hit its own monthly free-tier limit and return no data. That source is reported as degraded rather than silently empty.

Social removals are not emitted. Social profile detection is noisy, so a profile that disappears between runs is not reported as a removal. Only new profiles and significant follower swings are emitted.

The snapshot is latest-state only. It holds current state plus the most recent funding event keys, bounded. It is sized for next-run diffing, not as an archive.

previous_snapshot is single-company only. It is ignored in batch mode, where per-domain stored snapshots are used.

❓ FAQ

Why did my first run return no changes?

Because it was the baseline. It recorded the state to diff against. Run it again on your cadence and you will get deltas.

One source says degraded. Is the row still usable?

Yes. The other sources still produced events and the degraded source's prior state was carried forward, so nothing is falsely reported as removed. Just do not read an absence there as a change.

Does it scrape anything itself?

No. It orchestrates fleet sub-actors and normalizes their output into one change-event schema.

Do I have to store the snapshot?

No. The actor stores a per-domain snapshot itself. snapshot is returned in case you would rather hold it, and you can pass it back as previous_snapshot.

Why is a follower count change low severity?

Because acting on it is almost always wrong. Severity exists so a real buying signal is not buried under drift.

🧩 Want other GTM data?

Mamba Labs builds custom actors for B2B go-to-market teams. The public versions of that work live here on the Store, so our users get the same tooling we build under contract.

🧑‍💼 GTM Hiring Signal Scraper	🧱 Tech Stack Detector
📡 B2B Buying Signals Aggregator	🔑 Job Board Keyword Scanner
🔗 Domain to LinkedIn URL Resolver	🎯 ICP Fit Scorer
📋 Job Posting Monitor	📬 Domain Deliverability Checker
🏢 Company Firmographic Enricher	🌐 Company Social Presence Mapper
🪪 Company Identity Resolver	💰 Funding and Press Signal Scanner
👤 People Finder and Email Verifier	🚀 Prospect Engine
🤖 AI Tooling Detector	📮 Outbound Stack Detector
📝 Publishing Frequency Tracker	✉️ Work Email Waterfall Finder
⏩ Sequencer Lead Push	🏅 Workplace Program Detector
👥 Team Page People Extractor	🧭 Company Discovery List Builder

Every actor in the suite takes a domain or a company and returns one flat row, so they stack in the same Clay table without reshaping anything.

🛠️ Need something custom built for you or your team? Tell us what you are trying to find and we will build it. Talk to Mamba Labs.

🆘 Support

Something wrong, or a change event you did not expect? Open an issue on the Issues tab with the domain and the row, and we will look at it.

ℹ️ Sourcing and legal. This actor scrapes nothing directly. Every field comes from Mamba Labs fleet sub-actors, each of which reads public company data: job boards, company websites, public news and press, and public social pages. Each change event carries the Actor ID of the source that produced it, so any value is traceable. You are responsible for how you use the output.

Built by Mamba Labs.