TwinsAI
Signal-based selling

Signal-based selling: act on buying signals, not lists

Which buying signals are worth acting on, how to score and combine them, the play and the exact message for each one, how fast to respond, what goes wrong (false positives, creepy outreach, privacy) and how to prove it works.

CHECKED NOVEMBER 18, 2026· 13 MIN READ· PART OF GTM AGENTS
The short answer

Signal-based selling means reaching out when an account shows evidence it may be in market, instead of working a static list. Buying signals include first-party engagement, website visits, third-party intent data, job changes, hiring, funding and technographic changes. The method is to score fit and timing separately, act only when several signals stack, respond fast, and measure meetings per signal.

Key takeaways
  • Buyers decide early. In 6sense's 2025 survey of more than 4,000 B2B buyers, 94% of buying groups had ranked preferred vendors before first contact. Signals are how you get into that window.
  • Score fit and timing separately. Fit says whether an account could buy; signals say whether it might buy now. A strong signal at a poor-fit account is still a poor-fit account.
  • One signal is a hint; two or three stacking within a few weeks is a reason to call. Weight first-party signals (your site, your product, replies) above third-party intent.
  • Every signal decays. Respond to hot first-party signals the same day, and treat most third-party signals as stale after a few weeks.
  • Never mention the signal in a way that feels like surveillance, and check the privacy rules before you identify individual visitors. Measure meetings and pipeline per signal type, against a holdout.

Why signal-based selling works now

Most of a B2B purchase happens before a seller is invited in. In 6sense's 2025 Buyer Experience Report (a vendor survey of more than 4,000 buyers), the split between independent research and seller engagement moved from 70/30 to 60/40, 94% of buying groups ranked preferred vendors before first contact, and they bought from that preliminary favorite 77% of the time (6sense, November 2025).

That changes what outbound is for. A static list reaches accounts at random points in their buying cycle, mostly when they are not buying. Signals tell you which accounts are moving, so you can show up while the shortlist is still forming. Signal tooling is now standard: Bloomberry found that 43% of companies hiring a GTM engineer run a website visitor identification tool and 35% run an intent data or signal platform (Bloomberry).

The types of buying signals, ranked by how much to trust them

Not all intent signals are equal. First-party signals come from your own channels and are rarely wrong about who did what. Third-party signals are inferred from activity elsewhere and are noisier. Here is how the common ones compare.

Buying signals compared (our framework)
SignalWhere it comes fromWhat it suggestsTrustUseful for about
Pricing, demo or comparison page visitYour website (first-party)Active evaluationHigh if repeatedDays
Product usage spike or limit reachedYour product (first-party)Expansion or conversion readinessHighDays to weeks
Reply, event attendance, content downloadYour channels (first-party)Interest in the problemMedium to high1 to 2 weeks
New leader in a buying roleJob-change data, LinkedInNew budget owner reviewing the stackMedium to highFirst 90 days in role
Champion changed companiesJob-change tracking on your customers' usersA buyer who already trusts you, at a new accountHighFirst 90 days
Hiring for the team you sell toJob postingsThe team is growing or the problem is getting worseMediumWhile the role is open
Funding roundPress releases, funding databasesBudget and pressure to growLow to medium alone1 to 3 months
Technographic changeTech-detection dataAdopted or dropped a related toolMediumWeeks
Third-party topic intent surgeIntent data co-ops and publishersResearch on topics related to your categoryLow alone, useful in combination1 to 3 weeks

How website visitor identification works

Website visitor identification matches anonymous traffic to a company or, in some products, a person. Company-level ID typically uses reverse-IP lookup; person-level ID matches cookies, device IDs and IP addresses against an identity network. Coverage is partial. RB2B, for example, lists 15-20% coverage on its Free, Starter and Pro plans and 35-45% on Pro+ in its pricing table (its FAQ quotes higher averages that it says vary by traffic), with person-level ID limited to US traffic (RB2B pricing and FAQ, checked 2026-11-18). Dealfront counts each unique company identified in a month once, however many times it visits (Dealfront pricing, checked 2026-11-18).

Where third-party intent data comes from

Intent data providers watch research behavior across many sites. Bombora, the best-known source, runs a data co-op of publishers and brands with its own tag on thousands of B2B sites and tracks more than 25,300 intent topics (Bombora, checked 2026-11-18). The output is account-level: a company is researching a topic more than its usual baseline. It does not tell you who, or whether they are buying from you or writing a blog post about your category, which is why it works best as a tiebreaker rather than a trigger.

How to score and combine signals

Keep two scores. Fit (0 to 100) comes from firmographics and how closely an account resembles your closed-won customers; it changes slowly. Timing (0 to 100) comes from signals; it rises fast and decays. Act when both cross a line, and route by the combination.

Starting timing points (our framework; tune to your own data)
SignalPointsHalf-lifeCap
Pricing or demo page, 2+ visits in 7 days405 daysOnce per week
Champion joined the account4045 daysOnce
New VP or head of the buying team3045 daysOnce
Reply or meeting request from anyone at the account507 daysNone
Hiring 2+ roles in the team you sell to2030 daysOnce per quarter
Technographic change in a related tool1521 daysOnce per quarter
Funding round1030 daysOnce
Third-party topic intent surge1014 daysOnce per 2 weeks

Worked example

An account with fit 82 hires a new VP of Sales (30 points) and, nine days later, two people visit your pricing page (40 points). On the day of those visits, the 45-day half-life has decayed the VP points to about 26, so timing is roughly 66. With a rule of fit above 70 and timing above 60, the account goes to a rep today. The same pricing visits from an account with fit 40 go to a nurture track instead. Real numbers should come from your own history: pull 12 months of closed-won accounts and check which signals preceded the first meeting.

Routing by fit and timing
Timing highTiming low
Fit highRep or agent acts now, same daySteady low-touch outbound; watch for signals
Fit lowCheck the signal is real; often a student, job seeker or competitorLeave alone

Plays and example messages for each signal

A good signal message refers to the situation the signal implies, not the signal itself. Nobody wants to read "I saw you on our pricing page." Use these as starting points and rewrite them in your own voice.

Repeat pricing or comparison page visits

Email from the account owner, same day

Subject:Pricing for a team your size

Body:Hi Dana, teams of around 40 sellers usually ask us the same three pricing questions: how seats work, what is included, and what onboarding costs. I wrote up straight answers for each. Want me to send it over, or is a 15-minute call easier?

New leader in the buying role

LinkedIn note, then email, within the first two weeks in role

Rep:Congrats on the new role, Priya. Most sales leaders we talk to spend their first 90 days deciding what to keep in the stack and what to fix. If outbound connect rates are on that list, I have a one-page benchmark that might be useful. No pitch attached.

Champion moved to a new company

Call or personal email, within a week of the move

Rep:Hi Marcus, I saw you've moved to Northwind, congratulations. You knew exactly how we set things up at your last team. Once you've settled in, would it help to compare notes on what is working at Northwind today?

Hiring for the team you sell to

Email to the hiring manager

Subject:Ramping the four new SDRs

Body:Hi Sam, four open SDR roles usually means a ramp problem in about 60 days: new reps, empty calendars, and a manager listening to calls at night. Here is how other teams get new hires to their first meetings faster. Worth a look before they start?

Funding round

Funding alone is a weak signal: every vendor in your category sees the same announcement on the same day. Use it to raise priority on accounts that already fit and already show another signal, and lead with the plan the money implies (hiring, a new market) rather than the round itself.

Technographic change or third-party intent surge

Email, combined with a fit-qualified account

Body:Hi Lee, teams that move off a single-line dialer usually hit the same question within a month: how to keep connect rates up without burning their numbers. Here are the three checks we would run first. Happy to walk through them if it is on your list this quarter.

Response-time rules: how fast to act on each signal

Signals decay, and the slow response is the norm. In research on how US companies respond to web leads, published in Harvard Business Review, 37% of companies responded within an hour, 24% took more than 24 hours and 23% never responded at all (HBR, 2011). The data is old, but the pattern still shows up wherever teams measure it. More on inbound timing in speed to lead and the speed to lead definition.

Response-time targets by signal (our framework)
SignalFirst touch withinChannel
Inbound form or meeting requestMinutesPhone, then email
Repeat pricing or demo page visits from a fit accountSame business dayEmail or call from the owner
Product usage spike or plan limit reached1 business dayAccount owner, in-app or email
Champion job change1 weekPersonal email or call
New leader in role2 weeksLinkedIn, then email
Hiring, funding, technographic change1 to 2 weeksSequence, prioritized by fit
Third-party intent surge aloneDo not act aloneRaise priority only

Pitfalls: false positives, creepy outreach and privacy

  • False positives. Visitor ID often maps traffic to the wrong company (VPNs, ISPs, shared offices). Job seekers read careers and about pages; students and competitors read pricing pages. Exclude careers pages, known competitors and your own customers before scoring.
  • Over-alerting. If reps get 40 alerts a day, they act on none. Send only accounts that cross both thresholds, with the reason written out.
  • Creepy messaging. "I noticed you visited our site" tells the prospect they were tracked and gives them nothing. Write to the situation, never the surveillance.
  • Signal theater. Buying three signal tools without a play behind each one is the most common failure. Start with one or two signals you will act on every time.
  • Existing customers and open deals. Route signals from accounts with an owner or an open opportunity to that owner, not into a new sequence.

How to measure signal-based selling

Signal metrics worth tracking
MetricHow to calculateWhat it tells you
Meetings per 100 signaled accountsMeetings booked / accounts that crossed the threshold x 100, per signal typeWhich signals actually predict interest
Time to first touchMedian hours from signal to first outreachWhether the process is fast enough to matter
Pipeline per signal typeOpportunity value where that signal fired in the 30 days before the first meetingWhere to spend on data
False positive rateSignaled accounts disqualified on first contact / signaled accountsWhether filters and fit scoring work
Lift over holdoutMeeting rate of signaled accounts worked vs signaled accounts left aloneWhether the signal causes results or just correlates
  1. Pick two signalsStart with one first-party signal (pricing page visits) and one people signal (new leader or champion move).
  2. Hold out 10 to 20%Leave a random share of signaled accounts untouched for 60 days so you can measure lift, not just activity.
  3. Review weeklyLook at meetings per signal and false positives, then adjust points and thresholds.
  4. Add signals one at a timeOnly add a third-party source when the first two are running reliably and you can attribute what each adds.

Frequently asked questions

What is signal-based selling?

Signal-based selling is prospecting triggered by evidence that an account may be in market, such as repeat pricing page visits, a new leader in the buying role, a champion changing jobs, hiring, funding or a technographic change, instead of working a static list in order. It pairs a fit score with a timing score and acts when both are high.

What are buying signals?

Buying signals are observable events that suggest an account is evaluating or about to evaluate a purchase. First-party signals come from your own website, product and channels; third-party signals come from outside sources such as job-change data, job postings, funding news, technographic data and intent data co-ops.

What is the difference between intent data and intent signals?

Intent signals is the broad term for any evidence of buying interest, first- or third-party. Intent data usually means third-party data: providers such as Bombora observe research activity across a network of B2B sites and flag companies researching a topic more than their usual baseline. It is account-level and best used alongside other signals.

How accurate is website visitor identification?

It identifies a share of traffic, not all of it. RB2B, for example, lists 15-20% coverage on its standard plans and 35-45% on its premium plan in its pricing table, with person-level ID limited to US traffic. Matches can be wrong because of VPNs and shared networks, so confirm fit before acting.

Is website visitor identification legal?

It depends on where your visitors are and whether you identify companies or individuals. Under UK guidance, personal data about business contacts is covered by the UK GDPR, and tracking technologies fall under PECR. In the US, state privacy laws such as the CCPA apply. Get advice from your own counsel; this is not legal advice.

Sources
  1. 1.6sense, The Timeline for Influencing B2B Buyers Is Shrinking: 2025 Buyer Experience Report (vendor survey, 4,000+ buyers; checked 2026-11-18) (2025)
  2. 2.Bloomberry, I analyzed 1000 GTM Engineering jobs, here is what I learned (updated 2026-11-10; checked 2026-11-18) (2026)
  3. 3.Harvard Business Review, The Short Life of Online Sales Leads (Oldroyd, McElheran, Elkington) (2011)
  4. 4.RB2B, FAQ: coverage, US-only person-level ID, compliance (checked 2026-11-18) (2026)
  5. 5.Dealfront, Pricing: what counts as an identified company (checked 2026-11-18) (2026)
  6. 6.Bombora, Bombora's Data Co-op (checked 2026-11-18) (2026)
  7. 7.Information Commissioner's Office (ICO), Business-to-business marketing (2026)
  8. 8.Information Commissioner's Office (ICO), Direct marketing using electronic mail: what else do we need to consider? (tracking pixels) (2025)
  9. 9.State of California Department of Justice, California Consumer Privacy Act (CCPA) (2026)
Keep reading

Spend your reps' day on conversations, not ringing phones.

TwinsAI parallel-dials up to 10 lines at once, validates every number with the carrier before it rings, and bridges a live prospect to your rep in under 400ms. It works inside HubSpot, Salesforce, Close, Attio, Outreach, Zoho, HighLevel, Pipedrive, Apollo and SmartMoving, with no seat minimum.

Book a demo ↗