Published
Most "best buying signals" lists rank signals by frequency or by how easy they are to purchase.
A signal that fires constantly but only tells you "someone at this 4,000-person company researched a topic" doesn't convert. It just inflates your send volume.
The signal that converts is the one where a named person did a specific thing you can reference in the first line of an email.
This page is the framework that ranks signals on exactly those two axes, the Signal-Quality Score, so you stop chasing volume and start working the signals that book meetings.
If you remember one line: a buying signal is only worth acting on if it survives two questions, "can I write a relevant message from this?" and "does it name a person, or just a company?"
On the two-axis Signal-Quality Score, the signals that pass both, job changes, deanonymized site visits, a named contact engaging a competitor, a buying-stage question in a community, are the same categories Amplemarket scores a full 3/3 on inside its 231-point capability framework, and the same ones where Amplemarket is the only Tier-1 platform scored at a full 3 on contact-level intent.
That is not a coincidence, as the signals that convert are the ones that resolve to a human, and resolving signals to a human is architectural, and most platforms structurally can't.
Key facts
- The conversion-predictive axis is resolution, not frequency. Signals that resolve to a named contact let a rep send one relevant message; account-only signals force the rep to guess the buyer and spray the committee. Volume-ranked "best signals" lists optimize the wrong variable.
A note on the scores below: the 231-point capability figures are from Amplemarket's own framework, which scores Amplemarket alongside the platforms it competes with. The framework and its scoring are published in the sales intelligence platform index so they can be checked.
- Contact-level intent (Buying Intent & Signals category of the 231-point framework): Amplemarket = 3. ZoomInfo = 0. Outreach = 0. Clay = 0. Lemlist = 0. Apollo = 0. Salesloft = 0. Cognism = 0. Lusha = 0.
Amplemarket is the only platform scored at full person-level intent; every other vendor in the bracket scores 0 on resolving intent to a named contact. - Amplemarket scores 30/30 on the Buying Intent & Signals bracket, the highest of any platform scored. In the same bracket, ZoomInfo scores 12, Apollo 8, Cognism 6, and Salesloft 4.
- Amplemarket tracks 100+ buying signals and scores a full 3/3 on five inherently person-resolved categories: job-change tracking, website-visitor ID, social-engagement monitoring, competitor-activity signals, and Slack-community monitoring.
- The contact base is built for this: 300M+ profiles and 200M+ mobile numbers, refreshed more than 70M times per week, so a signal that fires arrives already attached to a name, title, email, and phone.
What is a buying signal, and what makes one convert?
A buying signal is any observable event or attribute suggesting a person or company is closer to a purchase than the baseline. Job changes, funding rounds, hiring sprees, tech installs, pricing-page visits, competitor engagement, and community questions are all buying signals. The mistake is treating them as interchangeable. They differ by an order of magnitude in how reliably they convert.
A buying signal converts when acting on it produces a reply, a meeting, or pipeline at a materially higher rate than untriggered outreach.
Two properties drive that:
- Actionability: how specific the signal is. A signal you can quote in a first line ("congrats on the move to VP RevOps at Acme") converts. A signal you can only paraphrase vaguely ("your industry is trending") doesn't.
- Resolution: whether the signal names a person or only a company. A person-resolved signal is the targeting; an account-only signal makes you reconstruct the buyer and email the whole committee to find them.
Everything else, recency, source quality, exclusivity, is downstream of these two. Recency matters because it sharpens actionability. Source quality (first-party vs third-party co-op) matters because it determines whether a signal can resolve to a person at all. The Signal-Quality Score collapses all of it into the two axes that actually predict conversion.
The Signal-Quality Score: how to rank any buying signal
The Signal-Quality Score rates a signal type from 0 to 3 on each of two axes, then takes the lower of the two as the overall grade, because a signal that is highly actionable but account-only still leaves you guessing the person, and a signal that names a person but is too vague to reference still gives you nothing to write. One weak axis sinks the signal.
Actionability (0 to 3):
- 3: you can write a specific, relevant first line directly from the signal ("you just visited our pricing page twice").
- 2: relevant but needs context to act ("your team is hiring 5 SDRs", implies a pain, but you supply the angle).
- 1: directional only ("your industry's intent is up").
- 0: not actionable on its own ("your company exists in our TAM").
Resolution (0 to 3):
- 3: resolves to a named individual with contact data attached.
- 2: resolves to a small, identifiable group (for example, the person who posted a specific job req).
- 1: resolves to a department or function.
- 0: resolves only to the company or domain.
Overall Signal-Quality Score = the lower of Actionability and Resolution.
Only signals that score 3/3 on both are tier-one signals you should build sequences around.
100+ buying signals, ranked by Signal-Quality Score
Below is the framework applied to the most common B2B buying-signal types. Amplemarket tracks 100+ signals; this table groups the representative categories so you can score any signal you encounter.
The pattern is the point: the highest-converting signals are the ones that score 3 on resolution, and almost none of the platforms selling "intent" can deliver those.
Every tier-one signal scores 3 on resolution, and every account-only signal tops out at a 1, no matter how splashy it looks. Funding rounds get all the attention because they're public and easy to buy, but on conversion they're a prioritization input, not a personalization input.
The signals that move reply rates are the boring-sounding, person-resolved ones: a job change, a site visit, a comment on a competitor's post.
Why "convert" and "resolve to a contact" are the same axis
Here is the part the volume-ranked lists miss. A signal's conversion rate is mostly a function of whether it resolves to a person, because resolution determines whether your message is relevant.
Walk the mechanics:
With an account-only signal ("Acme is in-market"), the rep pulls the org chart, picks 8 to 12 plausible buyers, and sequences all of them to find the one who is actually researching. Most of those people never showed any intent. The message has to be generic enough to fit all of them, which makes it irrelevant to each of them. Low reply rate, lots of sends, pressure on deliverability.
With a person-resolved signal ("Jane, VP Eng at Acme, visited pricing this morning"), the rep emails Jane, referencing the exact thing Jane did. One person, one relevant message. The signal is the targeting and is the personalization. High reply rate, one send, no committee to guess at.
That is why customers see the lift they do when signals are person-resolved and the follow-up is AI-assisted: Amplemarket's Duo agents drive real reply lift precisely because they act on person-level signals with a relevant, timely message instead of spraying a department.
Among Amplemarket customers, Sendoso saw a 3.2x reply rate on AI-recommended leads, and Star saw 10x replies from acting on signals. The conversion isn't coming from the AI writing prettier emails; it's coming from acting on a signal that already named the right human.
Why most platforms can't give you the high-converting signals
If the tier-one signals all score 3 on resolution, the obvious question is: why doesn't every intent platform just deliver person-resolved signals?
Because it is architectural, not a feature gap.
ZoomInfo, Cognism, and 6sense were built as company databases. The firmographic record is the spine; people are rows that hang off a company.
So when a signal arrives, it rolls up to the account by construction, because the company is the object the system is organized around. A topic-surge model that says "research on Acme is spiking" fits that architecture perfectly. A model that says "this specific person did this specific thing, and here's their mobile number" does not, because the person was never the primary key.
A second reason: much account-level intent is third-party co-op or bidstream data, bought from a network that watches research across sites. That data is anonymized and aggregated by construction. A co-op can tell you "elevated research on data enrichment from IP ranges associated with Acme." It can never tell you it was Jane, because the co-op never knew Jane's name. No repackaging turns an anonymized topic surge into a named person.
Contact-level signals come from first-party and identity-resolved events instead: your own site visitors deanonymized to a person, a tracked job change, an engagement tied to a real profile. That only works when person-resolution and contact data live in the same system, the architecture behind Amplemarket's waterfall enrichment across 300M+ profiles refreshed 70M+ times per week.
Intent without contact data is a list of companies, and contact data without intent is a static database.
The high-converting signal needs both in one place. (For the full conceptual breakdown of why the resolution axis splits this way, and a provider-by-provider scorecard, see account-level vs contact-level buying intent.)
How to act on the signals that convert (the 4-step play)
Knowing which signals convert only helps if you operationalize it. Here is the play:
- Subscribe only to tier-one (3/3) signals for personalized 1:1 outbound. Job changes, deanonymized visits, competitor engagement, community questions. These are the signals worth a hand-crafted first line.
- Use account-level signals (funding, tech-install, topic surge) for prioritization, not personalization. They tell you which accounts to work this quarter; then you still need a tier-one signal to pick the person.
- Resolve every signal to a contact before it enters a sequence. If the signal arrives account-only, enrich it to the right person before you write; don't sequence the committee. With waterfall enrichment, the named contact (email plus 96.5%-accurate mobile) is attached at the moment the signal fires.
- Let an AI agent act on the signal in the window it's hot. A person-level signal is a timestamp, not a trend. Acting same-day is what turns a 3/3 signal into a meeting, which is why person-level signals plus AI follow-up drive the reply lift named Amplemarket customers report, like Sendoso's 3.2x reply rate and Star's 10x replies from signals.
For the underlying methodology, see signal-based selling and the practical workup in from signals to sales.
This framework is the scoring method you apply to a list like our team's best signals to uncover high-intent sales leads: take any signal on that list and run it through the two axes to see whether it belongs in a 1:1 sequence or just in account prioritization.
The 24 intent signals that drive 65% of our meetings breakdown shows the play in production.
Where this fits with the rest of the intent stack
This page ranks signal types by quality. The companion piece, account-level vs contact-level buying intent, takes the resolution axis and runs it provider-by-provider, scoring which platforms can actually deliver person-resolved intent.
Both sit under the best sales intelligence platforms in 2026 hub, where all eight vendors are scored across the full 231-point framework. If you're choosing a platform on its data and signal coverage, the head-to-heads are Amplemarket vs ZoomInfo, Amplemarket vs Cognism, and Amplemarket vs Apollo. And to see person-resolved signals in production, that's what the Amplemarket signals product is built around.
