Buying Signals in B2B: Types and Examples

Six seats became eleven on one of your trial accounts this week. A different customer opened a support ticket asking whether you offer SSO, which you only sell on the enterprise plan. A third quietly downgraded in Stripe last night. Those are three buying signals, each one dated and specific, and none of them is on the account record your rep opens tomorrow morning.

Most of what gets written about signals covers the other kind: funding rounds, new hires and tech stack changes, bought from a feed and shared with every other vendor who subscribes to the same one. Those count, and we will get to them. The half nobody can sell you is the one you already own, sitting in billing, support and product tools that sales rarely opens.

This piece is a companion to our guide to lead scoring, which covers how fit and intent combine into one number. Here we stay on the intent side: which signals exist, where each one lives, and what should happen when one fires.

What buying signals mean in B2B

A buying signal is a change that makes a purchase more likely soon. The word doing the work is change. A 150-person SaaS company in your target industry is a fit. That same company posting its first RevOps job is a signal, because something moved and it has a date on it.

Most definitions skip that distinction, and it matters once you start scoring. Fit barely moves from one month to the next. Signals decay: a funding round from last week means something different from one eighteen months old, even though both fill the same field.

A quick test for any candidate signal:

• It has a timestamp. You can say when it happened, not just that it is true.

• It points at a problem you solve. A new CFO is a signal for a billing tool and trivia for a design tool.

• Someone can act on it this week. If nobody would change who they call, it is a report.

The vocabulary around all this is messy. "Sales signals" is the umbrella most reps use for everything above. Buyer intent signals usually means research behavior, like someone reading comparison pages or searching your category.

Intent data is narrower still. It is a commercial product: a vendor watches content consumption across a network of publisher sites and sells account-level topic surges. Signal-based selling is the practice of timing outreach to any of these. Marketing textbooks map the whole thing onto the stages of the buyer decision process, and honestly, we would not lose sleep over which stage a given signal belongs to.

Third-party buying signals: funding, hiring, job changes, tech stack

Third-party signals are public events about a company, detected by someone else and sold to you. Four categories carry most of the weight:

• Funding. New money usually turns into new headcount and new budget lines within two quarters. The window is weeks, not months.

• Hiring. A job post is the most honest document a company publishes. Three open SDR roles tell you what they are about to spend on before any press release does.

• Job changes. A champion who bought from you at one company and just started at another is the warmest cold account you will ever find.

• Tech stack. A company that adds a tool adjacent to yours, or drops the one yours replaces, has a switching moment you can see from outside.

All four share one weakness: everyone else can see them too. When a company announces a Series A, its founders get a wave of congratulations emails within 48 hours, most of them from vendors reading the same funding feed. The signal is real. Your timing advantage is gone by the time you read it.

Intent data sits in a slightly different spot. It is modeled from research activity rather than observed as a public event, and it arrives as a topic surge at the account level. We have seen it work for teams with long enterprise cycles and disappoint almost everyone else, though people who run ABM programs for a living probably have more firsthand data on the big intent co-ops than we do.

A tangent, since we are here. The job-change play has been run so hard that "congrats on the new role" reads as a template the second it lands, and we suspect it will be the first signal play to stop working outright. Third-party signals are still worth having. They just need company, and the best company they can keep is something only you know.

Buying signals examples: first-party signals your systems already record

Your billing system, support desk and product analytics log signals every day. They log them precisely, with timestamps, about accounts that already know your name. Almost none of it reaches the CRM, because each tool belongs to a different team: finance watches billing, support owns the tickets, and product owns the event stream.

A few of these as plays, the way we would actually run them.

The SSO ticket. A 40-seat customer on your mid plan asks support whether SSO is available. Support answers, links the pricing page, and marks the ticket solved.

That ticket is the clearest expansion signal you will get all quarter. The account owner never hears about it.

The second signup. A company went closed-lost in March over timing. In August, someone new at the same domain starts a trial. Your CRM shows the old deal and your product database shows the new user, but nothing joins the two, so the new trial gets the same onboarding emails as a stranger.

The quiet downgrade. Billing records a plan change the moment it happens, and Stripe's subscription webhook events fire for it. A downgrade after a usage drop is churn risk. A downgrade the same week a new finance lead joined is a budget review, and that call needs a different opener.

The usage spike. Weekly active users on a free workspace go from three to nine. Product-led teams call these product qualified leads, and that guide covers the usage side in depth. The narrower point here is that a spike says something about the account, not only the user, and it should move the account's score.

The first time we put our own support tickets next to our account list, the best lead of the month was a ticket that read "can we add more people?" It had been marked solved nine days earlier. Nobody in sales knew it existed.

What these signals have that no feed can match is exclusivity. A funding round goes to every vendor with the same subscription. The ticket about SSO went to you.

Stacking buying signals into an account score

One signal on its own is a coin flip with better odds. Three, from different sources, and the account is telling you something. The mechanics of turning inputs into a number are in the lead scoring guide above. What changes when the inputs are signals is time.

Rules we would start with:

• Score the account, not the person. The trial user, the person filing tickets and the new VP are often three different people at one company, and only the account sees all three.

• Decay every signal. Usage and tickets fade in days, funding in weeks, a job change over about a quarter.

• Let combinations outweigh sums. A raise plus a hiring push plus a usage spike describes a company building the exact thing you sell into, which is worth more than three separate points.

• Keep negative signals. A cancelled subscription or a month without logins should pull the score down, not just stop adding to it.

• Gate on fit first. A signal on a company you cannot serve is noise with a timestamp.

Here is how that reads on one account.

Northwind Analytics raised a Series A five months ago and has two open SDR roles. Inside your own systems, their workspace went from four active seats to eleven in three weeks, and last Tuesday their ops lead asked support about a Salesforce sync.

Individually, each of those is a maybe. Together they score 88, and the reason attached reads "funded, hiring SDRs, seats tripled, asked about Salesforce." A rep can check every clause of that in under a minute, which is why the reason matters as much as the number.

Doing this by hand works for twenty accounts. Past a few hundred, the ranking only means something if every account is compared against the others at once, which is the argument behind scoring every account against your ICP in a single pass.

Anyway.

Routing signals to your CRM and a sequencer

We once piped every signal into a Slack channel. After two months it held about 400 messages, and maybe six calls came out of it. Sales signals only pay off where the next action already happens, which means routing each kind somewhere specific:

• Score and reason onto the CRM record. A number reps sort by, and a sentence they read before dialing.

• New-logo accounts above your threshold into a sequence, where the first message opens with the signal instead of a line about their industry.

• Expansion signals from paying customers to the account owner, not the SDR queue. An SSO ticket from a customer is an upsell, and a cold sequence would be tone-deaf.

• Negative signals to whoever owns retention, before the renewal date rather than after it.

This loop is what we built Oneprofile around. Connect Stripe, Zendesk, PostHog, Gong or your Postgres database, and each account picks up its plan changes, tickets, usage and call mentions, matched on domain or email. Firmographics come from a waterfall of data providers ordered cheapest first, and a miss costs nothing. You can also search for companies a public signal just fired on, like a funding round or a hiring push, and they arrive already researched.

AI then scores every account against your ICP and writes the reason next to the score. Both sync back to your CRM two-way, and accounts above your line enrol in email and LinkedIn sequences whose opening line comes from the same research. Sources refresh on a schedule you set, so an account that upgrades on Thursday has moved before Friday's call list is built. The Team plan is $20 a month with every integration included, and the free tier renews 200 credits a month.

One thing to try before any tool. Open your support desk, search the last 90 days of tickets for "SSO", "seats", "API" and "invoice", and count how many came from accounts nobody in sales has spoken to this quarter. That count is your backlog of buying signals nobody acted on. Get started free when you would rather it stopped being a manual search.