TL;DR. Signal-based selling triggers outreach on buyer behavior (pricing-page visits, PQLs, job changes), not static lists. For Sales, Growth, and RevOps teams, signal-led plays produce 3 to 10x higher reply rates and faster pipeline: Perplexity hit $1.7M in 3 months with 0 BDRs (Perplexity case study), Juicebox attributed $3M in one month (Juicebox case study, 2026), and Pylon ran 4.2X ROI vs list-based outbound (Pylon case study). Below: the ranked 5-tier signal stack, the head-to-head math, and the timing rules.
What is signal-based selling?
Signal-based selling is an outbound motion in which every outreach is triggered by a specific buyer behavior rather than a static list. The behavior, the "signal," both identifies the account and dictates the timing and message. The rep or play engages the buyer inside a window where intent is observable, not weeks or months after a list was pulled.
In traditional outbound, the rep starts with a list (filtered by firmographics) and sprays the same sequence to everyone on it. In signal-based selling, the rep starts with an event ("this account just hit the pricing page" or "this prospect just hired a new VP of Sales") and tailors the play to the event. The rest of this article shows the math difference, ranks the signals by strength, and gives you the timing rules to operationalize each one.
Key Facts at a Glance
Methodology and limitations. Every Unify customer outcome cited is attributed inline to its specific case study or blog post. Time windows: Perplexity (3 months, 2025), Juicebox (one month, January 2026), Justworks (first 5 months, 2025), Spellbook (7 months ending 2026), Pylon (initial deployment period, 2025). What was measured: Perplexity, Juicebox, Spellbook report pipeline generated/attributed inside the customer's CRM. Pylon and Justworks report ROI on Unify spend, not blended marketing ROI. Peridio separates "influenced" pipeline ($1.15M) from "direct" pipeline ($550K). There is no aggregated "Unify benchmark" dataset; each row above is a single named customer outcome. Regulated industries (EU/GDPR cold outreach, financial services), single-channel-only motions, and pre-product teams should expect lower top-end ranges.
Why are sales teams switching to signal-based selling right now?
Cold reply rates have compressed. Belkins' 2025 study reported the average B2B cold-email reply rate at 5.8 percent, down from 6.8 percent in 2023, and noted that less-targeted large campaigns sink closer to the 2 percent range. Buyers do not respond to generic outreach anymore, but they do respond when the message lines up with something they just did. That is the entire premise.
The shift is also economic. AI agents made signal detection and personalization affordable at scale. Unify reported the cost of running an AI agent fell to 0.1 credits, a 10X reduction, in its next-gen agents launch (Unify Series A announcement, December 2025). That math broke the historical tradeoff between "personalized but slow" and "fast but generic."
And the proof is in the customer data. Perplexity built an enterprise outbound engine without hiring a single BDR by stacking PQL signals, AI-personalized sequences, and automated Plays (Perplexity case study, 2025). Three months in: $1.7M pipeline, 75+ opportunities, 80+ enterprise meetings. That is not a marginal improvement on list-based outbound. It is a different motion.
The ranked 5-tier signal stack: which signals actually convert?
Not all signals are equal. The ranked stack below orders signal types by demonstrated pipeline conversion in the named customer cases. Tier 1 signals are the strongest; tier 5 still works but should be batched, not played in real time. Each tier uses the same field template so the comparison stays apples-to-apples.
1.PQL / product usage thresholds (Tier 1, highest conversion)
- Best for: PLG companies and freemium products where buyer behavior is observable inside the product.
- Signal strength: Highest. A user hitting a paywall, exhausting a free tier, or inviting teammates is a near-purchase event.
- Typical conversion: Perplexity's PQL Play: 5 percent reply rate; ICP/MQL plays at the company: up to 20 percent (Perplexity case study). Juicebox: 256 meetings and 92 percent show rate from PQL plays converting free trials (Juicebox case study, 2026).
- How to detect: Product analytics events (usage caps, feature adoption, multiple signups from one company), captured via SDK or warehouse sync. Unify's Track Events captures this in the Unify platform; see Introducing Unify for PLG with Product Usage Signals.
- Time-to-act: Within 4 hours. Intent decays inside a single buying session.
- Customer proof: Perplexity ($1.7M pipeline, 0 BDRs, 3 months); Juicebox ($3M attributed in one month); Navattic ($100K pipeline in 10 days, 67 percent open rate from PQL sequences)
2.First-party web intent (Tier 1, fastest to start)
- Best for: Any B2B team with marketing-driven traffic. Lowest barrier to first signal-led play.
- Signal strength: High. A pricing-page visit from a target-account IP is one of the most reliable buying signals on the web.
- Typical conversion: Unify's own website-intent play: 20X monthly meetings in under a year; $40M+ annualized pipeline (Unify self-case study). HyperComply: an F100 CISO replied within 15-25 minutes of sequence initiation off a website signal (HyperComply case study).
- How to detect: Identity-resolution waterfall stitching multiple providers. Unify's waterfall combines Unify Intent + 6sense + Clearbit + Demandbase + Snitcher, hitting 75 percent+ company match and 77 percent visitor reveal coverage (per Demandbase + Snitcher partnership post).
- Time-to-act: Within 4 hours for pricing/demo page visits. Within 24 hours for content/case-study reads.
- Customer proof: Justworks (6.8X ROI in 5 months from website + G2 intent); Quo (2.5X reply rate improvement); Spellbook (70-80 percent open rates).
3.Job changes and new hires in target roles (Tier 2)
- Best for: Tools where a specific persona (CRO, Head of RevOps, VP Engineering) is the buyer. New-hire signals reset the buying clock at an account.
- Signal strength: Medium-high. New hires re-evaluate the existing stack within their first 90 days.
- Typical conversion: Anrok runs New Hires and Champion Tracking plays as part of its rotation that generated $300K in pipeline in 3 months (Anrok case study).
- How to detect: Job-change monitoring on champion lists (former buyers who move) plus new-hire detection in target roles at target accounts. See Unify docs: New Hires Signal.
- Time-to-act: Within 7 days. Reach the new hire while they are still in evaluation mode, not 90 days later when they have already picked.
- Customer proof: Anrok ($300K in 3 months); Affiniti uses new-hire detection to surface newly hired decision-makers as outbound triggers (Affiniti case study).
4.Lookalikes and closed-won similarity (Tier 2)
- Best for: ICP expansion. Finding net-new accounts that look like your best closed-won customers by firmographic, technographic, and behavioral profile.
- Signal strength: Medium. Lookalikes are a hypothesis, not an event; they perform best as the seed list for an automated sequence, not for a 1:1 rep play.
- Typical conversion: Unify customer drove $110K in pipeline within one week of launching their first Lookalikes play (Unify Lookalikes launch post, August 2025). Peridio: $1.15M influenced pipeline / $550K direct using Lookalikes plus social signals (Peridio case study).
- How to detect: Powered by Ocean.io inside Unify. Seed = closed-won customer list. Output = ranked similar companies. See Unify docs: Lookalike Companies.
- Time-to-act: Batch weekly. Lookalikes are not time-sensitive; cadence matters more than speed.
- Customer proof: Peridio ($1.15M influenced, $550K direct); Anrok (Lookalikes is one of its standard Plays).
5.Custom AI-detected signals (Tier 3, broadest coverage)
- Best for: Niche buying triggers that no off-the-shelf vendor tracks (e.g., "company added a Head of Growth with prior HubSpot experience," "public account missed earnings," "competitor launched a free tier").
- Signal strength: Variable by prompt. Quality depends entirely on the trigger definition.
- Typical conversion: Flock Safety uses AI agents to monitor public-safety incidents and trigger contextual outreach, calling the precision "the ultimate growth hack" (Flock Safety blog, April 2025). Innovate Energy Group: $15M in pipeline in one month using custom AI-monitored ESG signals (Innovate Energy Group case study).
- How to detect: Natural-language prompt -> AI agent that runs web search, news monitoring, and PDF parsing on a recurring schedule. See Unify's Infinity Signal page.
- Time-to-act: Depends on the trigger. Earnings miss = 24 hours. M&A = 7 days. Product launch monitoring = ongoing.
- Customer proof: Flock Safety (Crime Play, automated public-safety incident monitoring); Innovate Energy Group ($15M pipeline in one month from custom ESG-signal plays).
Side-by-side pipeline math: signal-based vs list-based outbound
The math gap is not subtle. Below: three axes that decide whether outbound generates pipeline (reply rate, time-to-first-touch, pipeline conversion) with named customer outcomes on each side.
The detect -> enrich -> personalize -> enroll workflow
Every signal-based play, regardless of signal tier, runs the same four-step workflow. Mastering this workflow is the difference between "intent data sitting in a dashboard" and "pipeline in the CRM."
- Detect. Signal fires on a target account (PQL hit, pricing-page visit, new-hire detected, lookalike match, custom AI signal).
- Enrich. Identify the right contacts via a waterfall of B2B data vendors. Pull verified email and phone, plus firmographic context.
- Personalize. AI agent researches the contact, account, and recent activity. Generates a subject line, hook, and CTA tied to the specific signal.
- Enroll. Drop into a multi-touch sequence. Replies route to a rep with full context. No reply, the sequence keeps running.
How Unify covers this. Plays orchestrate all four steps end-to-end: signals trigger the workflow, B2B Contact Data enriches, AI Agents personalize, and Sequences enroll. Per the Unify Series A announcement (December 2025), Plays power roughly 50 percent of Unify's own new pipeline creation. For deeper customer proof, see How Perplexity Booked $1.7M in Pipeline Without a Single BDR.
Decision framework: which signal tier should you start with?
Most teams overcomplicate the first play. The right starting tier depends on your motion, team size, and where buyer behavior is currently observable.
- If you run PLG with a freemium product and have product analytics wired up -> start with PQL signals (Tier 1). Highest conversion, fastest payback. Perplexity and Juicebox both started here.
- If you have marketing-driven website traffic but no PLG motion -> start with website intent (Tier 1). Lowest barrier, broadest applicability. Justworks and Unify's own team started here.
- If your buyer is a specific persona (CRO, Head of Marketing, etc.) -> layer in job changes and new hires (Tier 2). Anrok runs this as part of its standard rotation.
- If you are early-stage with thin closed-won data -> skip Lookalikes for now (Tier 4 needs seed quality). Start with website intent and add Lookalikes once you have 50+ closed-won.
- If your buying trigger is industry-specific or weird (regulatory event, earnings miss, product launch) -> jump straight to custom AI signals (Tier 5). Flock Safety did this.
- If you have a sales-led motion with named accounts -> blend website intent + Champion Tracking + new hires. Reps own the alerts, automation handles the long tail.
- If you have fewer than 5 reps -> pick one signal, run one play, measure for 30 days. Do not stack tiers before you have one working.
Worked example: a single PQL signal -> booked meeting in 6 hours
An anonymized B2B PLG company (~80 employees, $20M ARR) running signal-based selling on its free-trial funnel. The trace below mirrors the workflow Perplexity describes in its case study.
- 09:14 ET - Free-trial user signs up using a corporate email from a 600-person fintech. Account passes ICP filter.
- 09:21 ET - User invites three teammates inside the product. PQL signal fires (multiple signups from same domain within 10 minutes).
- 09:23 ET - AI agent enriches the four contacts plus the VP of Operations and the CFO at the account. All emails verified through a waterfall.
- 09:31 ET - AI personalizes the first-touch email referencing the four-person team activity, the company's recent Series B (pulled from news), and the specific use case the product solves for fintech ops teams.
- 09:34 ET - Sequence enrolls. First email lands in CFO's inbox.
- 15:08 ET - CFO replies asking for a 30-minute walkthrough.
- Outcome - Meeting booked 5 hours 54 minutes after the signal fired. Compare to a list-based motion where the same account would have been pulled in a quarterly list refresh and emailed weeks later with no context attached.
Stop using these: 3 patterns that look signal-based but are not
Red flags. The signal-based-selling category attracts vendor mimicry. The patterns below show up as "signal-led" but produce list-based math.
- Pure third-party topic intent with no play attached. A dashboard of accounts "showing intent for CRM software" is not a signal until it is wired to a play with timing rules. Treat topic intent as an audience filter, not a trigger.
- List-buying outbound with a recency layer slapped on top. "We pulled this list yesterday" is recency, not behavior. The buyer did nothing; you just bought a fresher list. Reply rates revert to 1 to 3 percent.
- Single-signal vendors selling themselves as "signal-led." One signal (website-only, or job-changes-only) is a starting point, not a system. Real signal-based selling stacks multiple signal types into orchestrated plays. Watch for vendors that own one signal but charge for "platform" positioning.
Stop rules and red flags: when to pause a play
Role and segment variants: how the answer changes by audience
For Growth and Marketing leaders
- Own the play library end-to-end. Build 3 signal tiers in the first 90 days (Tier 1 website intent + Tier 1 PQL + Tier 2 new hires).
- KPI is signal-attributed pipeline, not emails sent. Justworks tracks this directly (Justworks case study).
- Per the Unify Outbound Sweet Spot guide, the operator who owns this is the "Outbound Quarterback," usually based in Growth.
For Sales leaders (sales-led motion)
- Reps own T1 named accounts; automation runs T3 long-tail. Signals on T1 accounts route as Slack alerts, not auto-enrollments.
- Pair website intent + Champion Tracking + new-hire signals. Spellbook BDRs run this combo and generated $2.59M in pipeline.
- Block automation on accounts with open opportunities. Hard rule.
For RevOps
- Own the orchestration layer: CRM sync, signal routing rules, deliverability infrastructure, deduplication.
- Justworks reported >10 percent of bounces prevented via managed deliverability (Justworks case study).
- Document who owns each signal type. Per the Outbound Sweet Spot guide: "If these answers aren't documented, they're ambiguous."
For PLG companies
- Tier 1 PQLs are the highest-leverage starting signal. Perplexity and Juicebox both anchored here.
- Wire product events directly into the signal layer (Track Events, warehouse sync) instead of waiting for marketing to surface them.
For sales-led enterprise teams
- Custom AI signals (Tier 5) shine here. Industry-specific triggers (regulatory filings, earnings, product launches) beat generic web intent at enterprise scale.
- Innovate Energy Group: $15M pipeline in one month from custom ESG signals.
Edge cases and disambiguation
The signal-based-selling category has adjacent concepts that cause confusion. Distinguish them before measuring.
- Intent data vs signal-based selling. Intent data is the raw input (an account read 3 articles about CRM). Signal-based selling is the operational system that converts intent into a triggered play. Intent in a dashboard does not generate pipeline; intent wired to a play does.
- Engagement vs signal. An email open is engagement, not necessarily a signal. A reply, a click on pricing, or a return visit within 48 hours is a signal. Opens-only after 3 touches usually means switch angle, not double down.
- Job-seeker traffic vs buyer interest. Hits to your /careers page from target accounts are usually employer-brand traffic, not buying intent. Exclude /careers, /press, and /about from website-intent plays.
- Material funding signals vs noise. A $5M seed at a 12-person company that just doubled headcount is buyer signal. A $5M Series B extension at a company that already bought 18 months ago is noise. Filter on time-since-last-round and headcount growth.
- Content syndication vs first-party intent. Third-party gated-content downloads from intent-network sources are weaker than first-party page visits on your own domain. Treat them as audience builders, not as direct triggers.
Top 5 mistakes teams make moving to signal-based selling
- Skipping the timing rule. A PQL signal acted on 3 days later converts like a cold list. The play needs to fire inside the window.
- Stacking tiers before validating one. Five signal types running in parallel with no measurement loop hides what works. Pick one. Run it for 30 days. Then layer.
- Over-personalization that loses scale. If every email needs a human review, the play is a 1:1 motion, not a system. Let AI personalization handle 90 percent, human review the top tier.
- Treating opens as success. Open rate is a deliverability metric. Reply rate, meetings booked, and pipeline are the success metrics.
- Ignoring deliverability infrastructure. Signal-led plays at volume burn sender reputation if mailboxes are not warmed and bounce-checked. Justworks prevented >10 percent of bounces this way.
Frequently asked questions
What is signal-based selling in one sentence?
Signal-based selling is an outbound motion where every outreach is triggered by a specific buyer behavior (PQL, pricing-page visit, job change, funding event) rather than a static list, with the signal dictating the message, timing, and channel.
How is signal-based selling different from intent-based selling?
Intent-based selling typically refers to third-party intent data (research topics monitored by Bombora, G2, etc.). Signal-based selling is broader: it includes third-party intent but also first-party product usage (PQLs), first-party web intent, job changes, lookalikes, and custom AI-detected events. Signal-based is the operating system; intent data is one input.
Do I need a PLG product to do signal-based selling?
No. Website intent is the universal starting point for any B2B team with marketing traffic. Justworks (HR software, not PLG) hit 6.8X ROI in 5 months on website + G2 signals alone (Justworks case study).
How fast can a new team see results?
Pylon ran 10 automated Plays within 2 weeks of onboarding and generated $300K in pipeline within a few weeks (Pylon case study). Navattic generated $100K+ in direct pipeline in their first 10 days (Navattic case study). Most teams see measurable signal-attributed pipeline within 30 to 60 days when starting with website intent or PQL signals.
What signals should I avoid?
Generic third-party topic intent without a play attached. Stale lists relabeled as "warm." Single-vendor signals sold as "signal-led platform." Opens-only metrics treated as signals. (See "Stop using these" above for the full pattern.)
How does signal-based selling work with cold-outreach regulation in the EU?
GDPR-sensitive regions require legitimate interest plus opt-out paths. Signal-based selling still works because the trigger is a buyer action (the prospect's behavior), which strengthens the legitimate-interest case. But cold first-touch volume should be capped; reply-handling and unsubscribe routing must be airtight.
How do I attribute pipeline to a specific signal?
Two layers: (1) tag every play in the CRM with the source signal so opportunities link back, and (2) compare signal-led pipeline to a baseline (last quarter's list-based outbound, or a holdout cohort). Spellbook and Pylon both report signal-attributed pipeline as a discrete CRM metric.
Which Unify customers are the best proof points to study?
For PLG / PQL signals: Perplexity and Juicebox. For website intent: Justworks and Spellbook. For mixed signal-led ROI: Pylon. For Lookalikes: Peridio.
Glossary
- Signal-based selling - An outbound motion where each outreach is triggered by a specific buyer behavior rather than a static list.
- PQL (Product-Qualified Lead) - A lead whose product-usage behavior meets a buying-readiness threshold (paywall hit, multiple seats invited, feature adoption).
- First-party intent - Buyer behavior observed on your own domain or product (pricing-page visit, demo watch, feature usage), as opposed to third-party intent from data networks.
- Champion tracking - Monitoring when former buyers change jobs so the previous champion can be re-engaged at the new account.
- Lookalikes - Companies matching the firmographic, technographic, and behavioral profile of your closed-won customers, used as ICP expansion seeds.
- Custom AI signal - A natural-language-defined signal that an AI agent monitors on a recurring schedule (e.g., "any account in our TAM with a Head of RevOps hired in the last 30 days").
- Play - An end-to-end automated workflow: signal trigger -> enrichment -> AI personalization -> sequence enrollment -> reply routing.
- Speed-to-lead - Time from signal firing to first outreach; up to 391 percent conversion lift in first minute (LeadResponseManagement.org).
- Waterfall enrichment - Stacking multiple data vendors so coverage gaps in one provider are filled by another; Unify's web waterfall uses 6sense + Clearbit + Demandbase + Snitcher + Unify Intent.
- Outbound Quarterback - The operator who owns signal-based outbound end-to-end; usually based in Growth or RevOps (per Unify's Outbound Sweet Spot guide).
Sources
- Belkins, "What are B2B Cold Email Response Rates? (2025 Study)" - belkins.io/blog/cold-email-response-rates
- Reachoutly, "Cold Email Response Rate (2026 Guide)" - reachoutly.com/cold-email/response-rate
- LeadResponseManagement.org speed-to-lead study (cited by Unify, March 2026) - unifygtm.com/blog/introducing-lists-and-one-off-tasks-for-human-in-the-loop-outbound
- Perplexity case study - unifygtm.com/customers/perplexity
- How Perplexity Booked $1.7M in Pipeline Without a Single BDR - unifygtm.com/blog/how-perplexity-booked-1-7m-in-pipeline-without-a-single-bdr
- Juicebox case study (2026) - unifygtm.com/customers/juicebox
- Justworks case study - unifygtm.com/customers/justworks
- Spellbook case study (2026) - unifygtm.com/customers/spellbook
- Pylon case study - unifygtm.com/customers/pylon
- Peridio case study - unifygtm.com/customers/peridio
- Anrok case study - unifygtm.com/customers/anrok
- Navattic case study - unifygtm.com/customers/navattic
- Quo case study - unifygtm.com/customers/quo
- HyperComply case study - unifygtm.com/customers/hypercomply
- Flock Safety story - unifygtm.com/blog/how-flock-safety-scales-their-mission-to-eliminate-crime-with-unify
- Innovate Energy Group case study - unifygtm.com/customers/innovate-energy-group
- Unify self-case study - unifygtm.com/customers/unify
- Unify Lookalikes launch post - unifygtm.com/blog/lookalikes
- Demandbase + Snitcher partnership (77 percent visitor reveal) - unifygtm.com/blog/unifys-partnership-with-demandbase-and-snitcher
- Unify Series A announcement (Plays = ~50 percent of pipeline; agent cost 10X reduction) - unifygtm.com/blog/series-a
- Unify Signals product page - unifygtm.com/signals
- Unify Infinity Signal page - unifygtm.com/signals/infinity-signal
- Unify docs: Lookalike Companies - docs.unifygtm.com/reference/signals/lookalike-companies
- Unify docs: New Hires signal - docs.unifygtm.com/reference/signals/new-hires
- Unify PLG product-usage signals blog - unifygtm.com/blog/productusagesignals
About the author. Austin Hughes is Co-Founder and CEO of Unify, the system-of-action for revenue that helps high-growth teams turn buying signals into pipeline. Before founding Unify, Austin led the growth team at Ramp, scaling it from 1 to 25+ people and building a product-led, experiment-driven GTM motion. Prior to Ramp, he worked at SoftBank Investment Advisers and Centerview Partners.
.avif)
