Signal-based outbound selling: find, verify and use buying signals

What counts as a buying signal

A signal has three properties. It is an event, so it has a date. It is observable, so someone other than you can check it. And it raises the odds that the company needs what you sell soon. A company that matches your ICP is a fit. A company that matches your ICP and just opened three SDR roles is a fit with a signal.

Timing matters because buyers decide early. In the 6sense 2025 B2B Buyer Experience Report (nearly 4,000 buyers), buyers made first contact with sellers about 61 percent of the way through their buying process, and they chose a vendor from their Day One shortlist 95 percent of the time, up from 85 percent a year earlier (6sense). A vendor that writes after the shortlist is set starts behind. A message tied to a recent, relevant event is one way to get onto the list sooner.

The last column is our own rule of thumb from running campaigns, and no published benchmark backs it. Test every window here against your own replies.

Where to source each signal

The tool facts below come from each vendor's own pages, checked on 18 October 2026.

Hiring and leadership changes in Clay

Clay's documentation lists signals for new hires, job postings, job changes, promotions, news and fundraising, topic intent and web intent. The job postings signal watches target companies for newly opened roles. The new hires signal covers people who joined target companies within the last three months. Job changes and promotions read a person identifier such as a profile URL, and you supply the table of companies or people to watch. Signals run on Clay credits: depending on the signal you pay per result returned or per record checked, and Clay shows an estimated cost for the cadence you choose. They are available on every paid plan except Flex, and web intent needs a Growth or Enterprise plan (Clay University: Signals in Clay).

Funding in Crunchbase

Crunchbase says most of its data comes from venture partners and contributors, and that it does not rely on scraping the web: 4,000+ venture partners who submit monthly portfolio updates, and 600,000+ executives, entrepreneurs and investors who update over 100,000 profiles a month. It also says it tracks government filings and more than a thousand news publications, and that its own team verifies daily updates (Crunchbase Knowledge Center). Much of that data is contributed, so treat a funding record as a lead to confirm.

The primary check for a US private placement is the SEC Form D. An issuer must file it no later than 15 calendar days after the date of first sale, which the SEC defines as the date the first investor is irrevocably contractually committed to invest (SEC: Form D). Even the filing can trail the money by two weeks, and not every round produces one. We date a funding signal from whichever public record appeared first.

Technology installs in BuiltWith

BuiltWith says it covers more than 127,000 internet technologies. Its lead-generation page describes lists of sites using a given technology, filtered by category, location, traffic rank and technology spend, with exports to CSV and XLS. A feature it calls Actionable Insights is described as flagging companies that look about to buy, upgrade or switch technologies by reading changes in their stacks, such as trials, experimentation and rebuilds. BuiltWith also says it does not provide personal information for EU and California residents, so you still need a contact source for the people (BuiltWith: Lead generation, BuiltWith home page). A detection is a lead to check. Open the company's site and confirm the tool is in use before you write a line about it.

Intent and alerts in LinkedIn Sales Navigator

LinkedIn's plan comparison lists alerts for saved leads and accounts on Core, Advanced and Advanced Plus, and Buyer Intent signals on Advanced and Advanced Plus but not Core (LinkedIn Help: Sales Navigator account types and features). LinkedIn's feature overview describes alerts on decision makers at saved accounts, and recommended leads that include buyer intent signals based on your activity (LinkedIn Help: Sales Navigator). Clay's web intent signal covers similar ground from your own traffic: it identifies companies visiting your site, the pages they viewed and how they arrived. Intent data is inferred, so use it to rank accounts and keep it out of the first line. "I saw you visited our pricing page" reads as surveillance.

Verify before you write

Every signal that reaches an email goes through three checks. Find the primary page: the careers posting, the press release, the SEC Form D, the product page that shows the tool, the profile that shows the new title. Record its URL and date in the lead record, so the first line can be audited later. Then confirm it still applies. A job posting may be closed, a tool may be one of several, a new executive may already have left.

If you cannot find a primary page, drop the signal and write a segment-level first line instead. The cold email copy playbook explains why a line the reader can check beats a clever one.

In the lists we build, the signal and its source URL travel with the contact into the sequencing tool, so whoever writes the copy can see what a line is based on. That is the core of our list building and CRM cleanup work. One example is the 50K+ verified leads we built for a UK B2B SaaS company using Sales Navigator, Crunchbase and enrichment tools (case study).

Turn a signal into a first line and a timing rule

The first line states the event and where you saw it. Keep it to one sentence about one signal, with no guess about what the company will do next.

The timing rule below is a working default from our campaigns. Send the first email within five business days of spotting the signal. Use one signal per email. If an account shows two, lead with the fresher one and keep the other for email two. When a signal passes the window in the table above, retire it and fall back to a segment-level line. If an account you already contacted produces a new signal, wait at least 60 days from the last touch unless they replied.

The windows are short because a first line is only checkable while the event is still visible. A posting that closed last month gives the reader an easy reason to delete the email.

How signals and fit scoring combine

Fit answers who could ever buy. Signal answers who might buy now. Score them separately, because a single blended number hides the case that matters most: a company with a strong signal and a poor fit looks warm and absorbs sends.

In our work the fit cut does the heavy lifting. For Fema Tape we found 12,000 firms, kept 8,000 after AI scoring (anything below 50 out of 100 was cut), contacted 850 in month one and got 32 positive replies against a target of 15 (Fema Tape case study). For Rapid Lanyard we segmented targets scored 75 and above out of 100 (Rapid Lanyard case study). For Igrek, 9,000+ firms found became 1,800+ ideal ones after scoring (Igrek case study). The method for building that fit score is in the ICP playbook.

For the signal score, give each type points, multiply by recency, and cap the total at 100. An example set of weights (illustration): hiring for the function you serve 30, funding within 45 days 25, a competitor's tool detected 25, a relevant leadership change 20, topic or web intent 10. Halve a signal's points once it passes its window and zero it after twice the window.

Worked example (illustration). Company A scores 82 on fit and has a verified hiring signal (30) plus funding announced 20 days ago (25), so its signal score is 55 and it goes in this week's batch. Company B scores 44 on fit and has a signal score of 70. It is not contacted, whatever the signal says.

Measure which signals pay

Log the signal type on every contact, then compare replies and booked meetings by type, and ignore opens. Our planning benchmarks are 2 to 3 percent of sends replying, 15 to 50 percent of replies positive and 30 to 50 percent of positive replies booking, so a signal-led segment that sits at the low end of that range after a few hundred sends deserves a second look at the offer or the list before the signal.

Keep the comparison honest about volume. Telling a 2 percent reply rate from a 3 percent one takes about 3,800 sends per group by the standard two-proportion approximation at 5 percent significance and 80 percent power (our arithmetic). With a few hundred sends per signal type you can only see large gaps, so use signal tests to find big differences and leave fine-tuning for later. Once a signal type proves itself, it also decides channel order: the multi-channel sequence playbook shows how we add LinkedIn and phone steps for the top-scoring accounts.

Questions

What is a buying signal in B2B outbound?

A dated, checkable event that makes a purchase more likely soon, such as a job posting, a funding round, a technology install, a new leader or a topic-research spike. If you cannot name the source and the date, treat it as a guess.

How many signals should one email mention?

One. Lead with the freshest verified signal and keep any other for the follow-up. A first line that stacks three observations reads as a scraped profile.

How fresh does a signal need to be?

We use 14 days for content or intent, 30 for hiring, 45 for funding, 60 for tech installs and 90 for leadership changes. These are working defaults from our campaigns with no published benchmark behind them, so test them on your own replies.

Do I need Clay to run signal-based outbound?

No. Clay's documentation lists signal types, but Sales Navigator alerts, Crunchbase and BuiltWith each cover part of the ground, and a spreadsheet with a source URL and date column works at low volume. Tools mainly save time on monitoring and enrichment.