How a live graph of 1.17 billion professionals becomes precise, in-market audiences for paid acquisition, lead generation, and demand campaigns — the few signals that carry the most leverage, worked in depth, then the full catalogue of what else we can slice on.
Buying a list is buying a guess about who might be a fit. A signal is knowing what just changed about a company or a person — a new round of funding, a decision-maker who just took the seat, a team that suddenly started scaling. That difference is the whole game in paid acquisition: the same creative and the same offer convert several times better when they land on an audience that is in-market this week rather than a static segment that was true last year. This document is built like an analyst's brief: it leads with the five signals that move the needle most, shows the data and the evidence behind each, and then proves the breadth — the dozens of other axes the same engine can target on, on request. Every statistic here is real, pulled from the production dataset, with the date it was measured.
The reason timing wins is structural, and it is now well documented. By the time a buyer talks to a vendor, most of the decision is already made — they research, shortlist, and form a preference while they are still anonymous. The job of a campaign is to be present and credible during that window, not to arrive after it has closed.
The practical consequence: broad targeting wastes most of its budget on accounts that are simply not in-market, and the accounts that are never show up on a static list. Targeting by change fixes both ends. It is also where the measured return sits — account-based, signal-led programmes beat untargeted spend on ROI across every region Forrester studied.
None of this is exotic. It is the established discipline the industry calls buying signals, intent data, or signal-based selling. The question for any data provider is simply: which signals can you actually produce, how fresh are they, and do they resolve to a real person you can reach? That is what the rest of this brief answers.
The market has converged on a recognisable set of signal families. Naming them up front does two things: it lets us speak the same language a media buyer or RevOps lead already uses, and it lets us be precise about which half of the map our data is built for.
| Signal family | Examples | Our data builds it? |
|---|---|---|
| People & leadership change | Job change, new C-suite, promotion, departure | ● strong |
| Financial & corporate events | Funding round, IPO, M&A, budget cycle | ● strong |
| Hiring & growth | Hiring velocity, role-specific reqs, expansion | ● strong |
| Technology / technographic | Stack in use, incumbent tooling, migration | ◐ via the builder layer |
| Firmographic & fit | Vertical, size, geography, seniority | ● strong |
| Web / digital intent | Third-party topic surge, reverse-IP visitor ID | ○ not us |
The single most useful thing to know about this market: most of what is sold as "intent" today is third-party topic surge — a co-op that watches which companies are reading about your category — plus reverse-IP de-anonymisation of website visitors. Those are real, and they are not what we do. The signals that buyers trust most and that convert best outside of that category are job change and funding / hiring — and those are built from primary career history, which is exactly what we hold.
The universe those signals filter is large in every vertical a campaign tends to target — the count below is companies, before any signal narrows them to the in-market few.
We could list thirty signals. That is a dump, not an analysis — and it would bury the few that actually decide a campaign's economics. So this brief is ranked. Part 1 works the five signals that carry the most leverage in depth: each one independently predicts budget, intent, timing, or fit strongly enough to build a campaign around. Part 2 is the catalogue — the rest of the axes, with proof and a use for each, so you can see the full surface area without losing the signal in the noise.
Scored on what each of the five core signals predicts:
| Core signal | Budget | Intent | Timing | Fit | Reach |
|---|---|---|---|---|---|
| Fresh funding | ● | ◐ | ● | ◐ | ● |
| Hiring & growth | ◐ | ● | ● | ◐ | ● |
| Job change / new-in-role | ● | ● | ● | ◐ | ● |
| Technographic / developer | ○ | ◐ | ○ | ● | ● |
| Contactability | ○ | ○ | ○ | ○ | ● |
● strong predictor · ◐ partial · ○ not the point of this signal. Contactability scores only on reach by design — it is the layer that turns any of the other four into a person you can actually message.
One more thing worth knowing before the deep dives: signals do not all fire at the same rate. A funding round is rare and high-value; a hiring surge is common and high-frequency. A campaign usually wants both — a few high-intent triggers and a steady, large flow to keep the pipeline full.
The data
The funding layer carries the round name, the date, the amount, and the named backers, kept fresh through the most recent quarters. Because it is structured and dated, "who raised, at what stage, when" is a clean filter — and the shape of the whole market falls straight out of it.
The algorithm
Filter the funding layer by stage and a trailing date window, then join to the company's current
decision-makers so the audience is people, not an info@ address. The output is an
ad-platform audience or an outreach list of named buyers at companies that closed inside your window.
// recently-funded buyers, last 90 days, growth stages
GET outricher.com/api/v2/signals/funding
?stage=seed,series_a,series_b
&closed_after=2026-03-13
&role=founder,vp,head // resolve to decision-makersThe data
Every one of the 5.6B work positions carries a start and (where it applies) an end date. That turns headcount into a time series: per company, we can measure net momentum, and — more usefully — the shape of it. A team that just opened its first sales roles is entering its go-to-market buying cycle; a team adding engineers three quarters running is scaling infrastructure.
The algorithm
Compute trailing-window hire and exit counts per company from position start/end dates, then read the deltas: net headcount growth as a coarse "is this working" filter, and department-level ramps (first SDRs, a marketing team forming) as the sharper buying-intent trigger.
The data
Because we hold 5.6B positions across 1.17B people, a role change is observable across the entire professional graph — not just inside a list you uploaded. A current position with a recent start date is a "new in seat" buyer. The same data finds a person who just moved into a role your product is built for, anywhere it happens.
| Decision-maker | New role | Company | Started |
|---|---|---|---|
| R. AlRashed | Partner | (UAE) | Dec 2025 |
| A. Frentz | Partner | (France) | Dec 2025 |
| J. Brooke | Principal | (US) | Dec 2025 |
| A. Huang | Principal | (US) | Nov 2025 |
Illustrative slice of the "started a new senior role last quarter" cohort — the same query runs against any title and any vertical.
The algorithm
Scan positions for fresh start dates at the seniorities and functions you target, and — for champion-tracking — diff a known contact list against current employers to catch moves into ICP-fit companies. No seed list is required for the first; the second works even on people you never had in a CRM.
The data
Most technographic vendors read the website layer — the scripts and tags on a company's marketing site. We can read the builder layer instead: 204M developer profiles with public open-source activity — the languages, frameworks, and tools the engineers actually use and contribute to, plus their reach and whether they are open to opportunities. For dev-tools, infrastructure, API, and security products, that is an audience the website-reading tools are blind to.
The algorithm
Segment the developer graph by stack (language / framework / tooling), by public output and reach, and by whether they signal openness — then resolve to a reachable identity. Pointed at a company, it profiles the tools its engineers really run; pointed at a project, it ranks the contributors into seniority bands.
The data
237M records are already enriched with contact data, backed by a deep email waterfall. That layer is what grounds every other signal in a person you can message — the gap where pure-intent and technographic vendors stop, because identifying an account is not the same as being able to reach the buyer inside it.
The algorithm
Resolve the signal — an account or a person — to an identity, then to the contact channels: personal email, phone, and social handles. The pre-enriched layer is checked first and costs nothing; live enrichment fills the gaps on demand, one inexpensive call at a time.
// one call resolves a target to a full, reachable dossier
GET outricher.com/api/v2/profile?profile=<target>
// → name · role · personal_emails · phone · socials · work historyThe five above carry most of the leverage. But the same engine slices the graph on many more axes — here is the breadth, grouped the way the market thinks about it, with the proof and a use for each. The point of this section is range: whatever a specific campaign needs to target on, the surface area is here.
A new C-suite or VP hire arrives with a ~90-day mandate to re-evaluate the stack — the same position-timing engine, filtered to senior titles.
proof · current senior role + fresh start dateuse · reach the new buyer before vendors are chosenNot just "a company is hiring" — which function is scaling. A doubling sales or marketing team is a far sharper buying trigger than headline headcount.
proof · function-level hire counts from 5.6B positionsuse · target the exact team about to buyA company's employees appearing in a new country signals market entry — and a need for local tools, services, and partners.
proof · employee location shift; geo down to cityuse · localised offers timed to expansionWhen a senior leader's role flips to "past," you know the week, the name, and where they went — a window for both displacement and recruiting offers.
proof · position end-dating with destinationuse · displacement & transition campaignsTarget by industry at scale — the universe is large and clean in every commercial vertical, from 273K IT-services firms to 204K marketing & advertising firms.
proof · indexed industry counts (chart above)use · vertical-specific creative & listsSlice SMB / mid-market / enterprise by headcount — including the depth to map 300,000+ employees at a single large company when needed.
proof · employee-count segmentationuse · match offer to company scaleResolve the actual humans in a target function — and reconstruct a buying committee from career data, for multithreading rather than a single contact.
proof · title + seniority across the graphuse · multithread the real decision unitTime-in-role and prior employer — target the recently-settled, or the alumni of a specific company (ex-unicorn, ex-incumbent) as a quality proxy.
proof · position dates + employer historyuse · lookalike-by-pedigree targeting237M records carry contact data; for senior decision-makers a third are reachable for free before any spend — the layer that makes every other signal actionable.
proof · 237M enriched · 35% senior freeuse · turn any audience into a reachable oneFollower counts and handles surface the people with audience — useful for influencer, affiliate, and creator-led acquisition, not just direct outreach.
proof · follower counts (six-figure reach observed)use · influencer / affiliate sourcingDeveloper and profile openness flags identify people signalling availability — the core audience for recruiting-led lead arbitrage.
proof · openness flags on the developer graphuse · recruiting & talent campaignsTarget down from country to region to city, and segment by language — so creative and compliance match the market, not just the continent.
proof · country→city geo; language segmentationuse · localised, compliant targetingThe real edge is layering: "raised in 90 days" and "hiring sales" and "new VP" is a far smaller, far hotter audience than any one alone. Each added signal compounds intent.
proof · signals combine in one query engineuse · build the smallest, hottest audienceCold outreach replies at roughly 1–3%; signal-led outreach runs many times higher. Stacking is how you trade a huge cold list for a small audience that actually answers.
proof · ~1–3% cold baseline (Instantly, 2026)use · spend less, convert moreThe pricing mirrors how a campaign actually runs: discovery and sizing are free, segmentation is free, and you pay only to put contact details on the people you decide to reach.
| Operation | Covers | Cost |
|---|---|---|
| Discovery · sizing · ranking · segmentation | Building every audience and applying every signal | Free |
| Pre-enriched contact (≈20% of the graph; 35% of senior buyers) | Personal email already on file | Free |
| Live profile enrichment | Full contact dossier, on demand | ~$0.02 |
A weekly, signal-driven audience of thousands of in-market companies costs effectively nothing to build. You spend only to reach the few hundred a campaign actually activates — which is the unit economics of an acquisition engine that never runs out of well-timed, reachable targets.
The takeaway: the value is not the 1.17B profiles — anyone can buy rows. It is the ability to query the professional economy as a live graph, so a static list becomes a timed, ranked, reachable audience: who just got the budget, who just took the seat, who is scaling right now, which engineers run the stack — each resolved to a person you can message. Lead with the five signals that decide a campaign's economics; reach for the catalogue when a specific play needs it.