Signal Types
Every signal BulkData.ai tracks.
Eleven signal types across people movement, record lifecycle, company events and technology — what fires each one, how fresh it is, and how it reaches the place your team already works.
What a signal is
A signal is a dated, verified fact about a company that changes how worth contacting that company is right now. It is not a lead list, and it is not a score on its own — it is the event underneath the score, with its source and its detection date attached, so a rep can open with the actual reason they are calling rather than an invented one.
Something happens
A company posts twelve new roles, closes a round, swaps a tool, or hires a VP.
We detect and verify it
The event is picked up from its source, checked against a second one where one exists, and attached to the account with a detection date.
It reaches you
Scored against your ICP and delivered to your feed, your inbox, your endpoint or your CRM.
Every signal, in one filterable place
The same eleven signal types the catalogue below documents are the filters you work with in the product — tick the ones your team acts on and the feed narrows to them.
The signal catalogue
All eleven signal types the platform emits, grouped for reading. Compare triggers, freshness and real examples to choose the events that matter to your team.
People movement
The four events that change who your contact is, where they sit, or whether they still work there — the category that decays a CRM fastest if nobody is watching it.
| Signal | What triggers it | Typical freshness | Example |
|---|---|---|---|
| Job change | A tracked contact turns up at a different company. A champion who liked your product at their last employer is the warmest first line you will get. | Typically within a day of the public profile change | Emma Whitfield left Vantify for Northbeam as Head of Ops. |
| Role change | An existing contact moves function inside the same company — useful for keeping your CRM’s champion and decision-maker fields honest. | Typically within a day of the public profile change | Emma Roberts moved from Marketing to Revenue Operations. |
| Promotion | A contact moves up a level in the same organisation. Seniority is usually the difference between a contact who can recommend and one who can sign. | Typically within a day of the public profile change | Emma Roberts promoted to Head of Revenue. |
| Relocated | A contact’s location changes — a different city, region or country. It re-routes territory ownership and it changes what time zone to call in. | Typically within a day of the public profile change | Daniel Okafor relocated from Manchester to Dubai. |
Record lifecycle
What is entering the index, what is leaving it, and what changed on a record you already hold. These are signals about the data itself rather than about an event at the company.
| Signal | What triggers it | Typical freshness | Example |
|---|---|---|---|
| Newly listed | A person or company enters the index for the first time — a company that has just started publishing, or a contact who has just become visible. | Appears on the crawl that first sees the record | Harborline Analytics entered the index this week. |
| De-listed | A record leaves the index — the company stopped trading, the profile was withdrawn, or the contact exercised a removal right. Worth knowing before you email it. | Applied on the crawl that first misses the record | Corely Ltd was de-listed after ceasing to trade. |
| Contact details | A verified email or phone number on an existing record changes, is added, or fails re-verification. The single most common reason a working list stops working. | Applied on the record’s re-verification cycle | A new verified direct dial replaced a dead one at Fenlow. |
Company events
Something changed at the account rather than on a person — the two events most reliably followed by new headcount and new tooling.
| Signal | What triggers it | Typical freshness | Example |
|---|---|---|---|
| Funding | A new round is announced or filed — seed through late stage, including extensions and debt facilities. | Typically within hours of the announcement or filing | Nexus SaaS raised a $12M Series A this week. |
| Company growth | Sustained expansion at the account — headcount climbing, offices opening, roles being opened across functions. Scaling operations, not just spending. | Aggregated and refreshed daily | Orbit Tech grew headcount 24% and opened an Austin office. |
Technology
What an account runs — and what it has just stopped running — tells you whether you are a replacement, a complement, or too early.
| Signal | What triggers it | Typical freshness | Example |
|---|---|---|---|
| Tech change | A tool in or adjacent to your category is detected on the account’s public web properties for the first time, or a previously detected one disappears from them. | Detected on a recurring crawl, usually within days | Harborline dropped its Marketo tag and added HubSpot. |
Other
A real facet in the product, not a placeholder: a detected, dated change that does not fall into the ten above still reaches your feed rather than being dropped.
| Signal | What triggers it | Typical freshness | Example |
|---|---|---|---|
| Other | Any verified change on a tracked company or contact that does not map to one of the ten named types — a registry update, a rebrand, a status change. | Follows the source’s own publication cycle | Brightbase filed accounts and changed registered address. |
How signals reach you
Four routes out of the platform. Pick one or all four — the filters you set apply to every one of them.
In-app feed
A live, filterable stream of every signal on the accounts and saved searches you follow, newest first.
Email digest
The same feed, batched into a daily or weekly email so nobody has to remember to open a tab.
Outbound webhook
A signed POST to your own endpoint the moment a signal fires, so your service decides what happens next.
See the payload shapeCRM push
The signal, the account and the matching contacts written straight into your CRM or sales-engagement tool.
Browse the integrations
How fit scoring works
A signal on its own only says something happened. The score says whether it happened somewhere you sell.
You define the ideal customer profile — the firmographics, the titles and the territories you actually sell to. Every signal that fires is then scored against that definition rather than against a universal notion of a good company, which is why the same funding round can rank high for one workspace and be filtered out of another. Scores move as your ICP changes, and as your team marks accounts as worked or dismissed.
What goes into a score:
- How closely the account matches the firmographics in your ICP — industry, size, region, business model
- Whether the people the signal concerns hold the titles and seniorities you sell to
- How recent the event is, and whether the same account has fired other signals around it
- The signal type itself, since some events matter far more to your category than others
- Your own history: the accounts your team has worked, won and explicitly dismissed
We do not publish the weighting between those inputs, and we would be inventing precision if we did — the relative importance is tuned per workspace as the model sees which accounts your team engages. Treat a score as a ranking of what to look at first, not as a probability of closing.
Signal questions
How fast signals arrive, where they come from, what they cost, and what you can do with one.
How quickly does a signal appear after the event?
It depends on how the underlying source publishes. Signals drawn from sources that publish continuously — professional profiles, job boards, funding databases — are typically detected and verified within hours of the source going live. Signals that depend on a source refreshed on its own daily cycle arrive on that cycle. Each signal type in the catalogue above carries the freshness we expect for it; treat those as typical behaviour rather than a service-level guarantee.
What sources are signals derived from?
Public and licensed sources: public professional-profile changes, funding and corporate-registry records, company career pages and hiring activity, and technology detection on public web properties. Every signal is verified against more than one source where a second source exists, and the record it attaches to carries the detection date. The exact source mix for a given signal type changes as sources are added or retired, so the catalogue describes what triggers a signal rather than naming a fixed vendor list.
Can I filter to only the signals I care about?
Yes, and most teams should. You can restrict a feed by signal type, by the account lists or saved searches you already keep, by firmographics such as industry, size and region, and by a minimum fit score. Filters apply to the in-app feed, to digests and to webhook subscriptions, so an alert you have filtered out of the feed will not reach your endpoint either.
Do signals cost credits?
Seeing a signal does not. Browsing and filtering the feed is free in the same way that searching and filtering records is free — credits are spent when you reveal a contact or export data. So acting on a signal by revealing the decision-makers at that account costs credits; watching the account and reading the signal does not. Which plans include buying signals is set out on the pricing section of the homepage.
How far back does signal history go?
Each account carries a rolling history of the signals detected on it, so you can see whether this month’s company growth follows a funding round from earlier in the year. The length of that window depends on your plan and is still being finalised, so read the date range shown on the account’s signal timeline rather than assuming a fixed number of months.
Can I trigger a sequence from a signal?
Yes. A signal can push the matching records into a CRM or sales-engagement tool through an integration, or fire an outbound webhook at your own service so you can start whatever sequence you already run. Teams without a native connector usually route the webhook through Zapier or Make. The signal carries the account, the signal type and the fit score in its payload, so the sequence can branch on any of them.
STOP SETTLING FOR BLIND SPOTS
Start selling to the marketsyou've been missing.
Search people and companies, enrich records with verified emails and phones, and move clean data into the tools your team already uses.
