The form-fill is dying as a discovery mechanism, and most B2B teams have not yet internalized what replaces it. The evidence for the death is not anecdotal: Salesforce's State of Marketing research finds 54 percent of marketers cannot determine lead quality even after forms are filled — the form tells you someone arrived, not whether anyone is buying (Salesforce via Scrap.io 2026). And Gartner's buying-journey research explains why: the average B2B purchase now spans more than 4.6 months and crosses seven channels before a decision, meaning most of the buying process happens before, between, and around the moments a seller can see (Gartner). The replacement for the form-fill is a stack of intent signals — observable behaviors that reveal active buying before anyone volunteers their email. This piece catalogs the seven signals worth instrumenting, where each fires in the journey, and how to triage them into a working queue.
Before the catalog, the principle: a signal is only useful if it is rare enough to mean something. Every company visits your website; almost none read three pieces on the same solution theme in a fortnight. The seven signals below are ranked roughly by that rarity logic — from broadly available to narrowly decisive — and the strongest discovery programs stack them, because two weak signals arriving together are often one strong one.
Signal one: first-party content clustering. When a single account consumes three or more pieces on the same solution theme within fourteen days, that is a research sprint, not browsing. The pattern is visible in any analytics stack that resolves visits to accounts, and it fires early in the 4.6-month journey — often before the buyer has shortlisted anyone. The nuance is thematic specificity: two pricing pages and a comparison guide cluster differently than three scattered blog posts. Instrument the cluster, not the pageview.
Signal two: AI-assistant referral traffic. The fastest-growing acquisition source of 2026 is referrals from ChatGPT, Perplexity, Claude, and Gemini (Digital Applied 2026) — buyers are asking AI assistants to shortlist and compare vendors. When your site starts receiving AI-referred visits from an account, two things are true simultaneously: the buyer is mid-research, and your name survived one round of algorithmic selection. Appearing in AI answers is therefore not just brand hygiene; the referral itself is an intent surface, and it is the newest signal on this list, which also means your competitors are least likely to be tracking it.
Signal three: job postings that name the problem role. Hiring is budget made visible. A posting for the exact function your product serves — the first revops hire, the procurement analyst, the demand-gen manager — means the org has allocated headcount to the problem, and headcount precedes tooling decisions by weeks to months. The signal is public, structured, and underused because it lives in a system (job boards) that sales teams do not habitually monitor. Wire your ICP's role keywords to a posting feed and treat each match as a 90-day window.
Signal four: tech-install and stack changes. New technology appearing at an account — visible in public telemetry, integration marketplaces, and job-posting requirements — predicts adjacent purchases. A company that just installed your category's favorite integration is re-platforming; a company that just dropped your competitor's favorite tool has a vacancy in its stack. This signal fires mid-journey, closer to evaluation, and pairs naturally with signal-based outreach that references the change without surveillance overtones.
Signal five: executive movement and funding events. New leadership resets vendor relationships — incoming executives replace inherited stacks at rates that make the first 90 days of a CRO or CIO tenure the single richest prospecting window in enterprise sales. Funding rounds work the same way on a different clock: fresh capital carries board-mandated spending plans, and the interval between a Series C and the tooling purchases it funds is measured in months. Both events are public, timestamped, and rich enough to segment by.
Signal six: community and practitioner Q&A activity. Buyers ask questions in niche communities — Slack groups, forums, subreddits — that they would never put in a vendor form, because communities are where it is safe to admit a problem. Activity there is attributable more often than teams assume: handles, signatures, and context clues resolve to accounts with modest effort. The signal fires earliest of all seven — sometimes in the problem-recognition phase months before any vendor contact — which makes it a compounding channel for teams patient enough to be useful where the questions are asked.
Signal seven: competitor engagement patterns. Support complaints, renewal-window chatter, and review-site activity around an incumbent are buying signals wearing someone else's jersey. A cluster of negative reviews at a competitor's flagship account, a public RFP that resembles an exit, a support thread going unanswered — each marks an account whose switching cost has just dropped. This signal fires latest and hottest, closest to the evaluation phase, and it rewards teams that maintain named-account vigilance on their category's incumbents.
A vignette per signal makes the texture concrete. Content clustering: a 400-person logistics software account reads your routing-optimization guide, then the API-pricing page, then a case study in their vertical, all inside nine days — that account is writing requirements somewhere. AI-referral: analytics show visits from a manufacturing group whose referrer is Perplexity, landing on your "alternatives" page — someone asked an assistant for options and you made the shortlist. Hiring: a mid-market fintech posts two procurement-analyst roles in one week; procurement tooling follows procurement headcount with near-mechanical reliability. Tech-install: an account's careers page starts requiring experience with your category's core platform — they are standardizing on an adjacent stack that your product integrates with. Executive: a new CRO arrives at a legacy-industry target; her first ninety days will include the vendor review your last four touchpoints could not get into. Community: a semi-anonymous question in a niche Slack about migrating data off an incumbent names the exact integration pain your product removes; the handle's bio resolves to an account. Competitor: three one-star reviews appear on your largest competitor's listing in a month, all citing the support gap your onboarding explicitly solves. None of these arrived as form-fills. All of them are buying.
Signal hygiene deserves its own discipline, because every signal type produces false positives, and an unmanaged false-positive rate burns the channel's credibility inside your own team. Content clusters misfire when a competitor, a journalist, or a student is doing the reading. Job postings misfire when roles are posted perennially or filled internally. Funding signals misfire when the round is debt, not growth capital. The corrections are structural rather than heroic: require two independent signals before outbound on a cold-fit account; discount recurring-poster roles after thirty days; and log every fired signal against its eventual outcome for a quarter, so each signal type earns a live conversion rate — and a calibration — instead of an assumed one. The goal is not zero false positives; it is a known false-positive rate the outbound team can price into its sequences.
Build the stack incrementally, in the order that pays for itself. Month one: first-party content clustering and hiring triggers — both run on data you already own or can subscribe to cheaply, and both produce their first routed accounts within weeks. Month two: AI-referral tracking, which needs only an analytics referrer filter and an account-resolution layer, and community monitoring for the two communities where your buyer is most concentrated. Month three: tech-install telemetry and competitor-watch on your five most valuable incumbent accounts. Executive-movement alerts slot in whenever, since the data is public and the workflow is episodic. Each addition should clear a payback bar the previous additions set — if month one's signals are converting, month two's budget is self-arguing.
Now the triage SOP, because seven signals without a queue is a notification storm. Score every fired signal on three axes: recency — fired within 14 days versus 30; specificity — names your category versus adjacent problem; and account fit — in-ICP versus borderline. AI lead scoring applied to that stack cuts wasted outbound on low-quality leads by 40 percent (Optifai 2026), and the operational rule that follows: accounts scoring above threshold route to signal-based outbound within 24 hours, mid-tier accounts enter a nurture track keyed to the signal's theme, and low-tier signals simply update the account map for context. The 24-hour rule matters because signals decay — a research sprint detected three weeks late is a courtesy call, not an interception.
The stack conclusion: signals find the account, verified contacts find the person, and account mapping holds the system together. A signal without a reachable contact dies in research; a contact without signal context dies in the inbox. The teams that will win discovery in 2026 are not the ones with the most signals but the ones with the shortest loop from signal to relevant touch — and that loop is an architecture, not a tool purchase. Stand up three signals this month — content clustering, hiring triggers, and AI-referral tracking are the fastest to instrument — and let the loop's conversion math argue for the rest.