Signals · AI GTM

Signal-based GTM

Your buyers are telling you they're in market - through your product, your website, their hiring, their stack. I design and build the infrastructure that turns those signals into pipeline.

Sound familiar?

  • You're paying for third-party intent data your reps ignore.
  • You've heard of Clay, but you don't really know if it's worth investing in.
  • Your product generates thousands of behavioural events a day that never reach the commercial team.
  • Outbound reply rates keep falling, so the answer keeps being "more sequences".
  • The lead score in your CRM can't explain itself, so your team doesn't trust it.

This is what it looks like when a revenue team has signals but no signal system.

Signals are an infrastructure problem

Vendors will sell you "signals" - job changes, funding rounds, website intent, product usage. The list is table stakes: your competitors can buy the same feeds. What separates teams using signals to create pipeline are 3 design decisions:

01

One signal is noise

A funding announcement means dozens of vendors email the same CFO by lunchtime. A single signal is never sufficient justification for outreach. Systems act on overlap - 2 or more validated signals on the same account.

02

First-party beats rented

Third-party intent tells you someone read an article. Your product tells you someone used a feature 12 times this week. The data you already own is more predictive than anything you can rent - it just hasn't been engineered into your GTM motion.

03

Score on your pipeline

Vendor intent scores are trained on thousands of other companies' aggregate data. They know nothing about your ICP, your sales cycle, or what preceded your last 50 closed-won deals. Signal weighting has to be trained on your revenue history - or the score is decorative.

What I build

A working system, shipped in sprints and measured on pipeline created, with a named owner.

1Audit
2Capture
3Score
4Act
5Measure
Signal audit

An inventory of the signals your business already generates - product events, website behaviour, CRM history, support conversations - mapped against the feeds you're paying for. The usual discovery: you're renting a worse version of data you already own.

Capture layer

The pipes: product usage, website intent, enrichment and third-party feeds flowing into one place, resolved to accounts and people. Built on your stack - warehouse, CRM, and tools like Clay where they fit.

Scoring model

Signal weights trained on your closed-won and closed-lost history. Every score explains itself - which signals fired, when, and why the combination matters - so reps act on it.

Plays, wired in

Each qualifying signal combination maps to a play: who reaches out, with what message, in which channel, inside what window. Routed into your CRM and sequencing tools automatically.

Measurement

Reporting on pipeline created per signal and per play. Weights get retrained as deals close, so the model compounds.

Team enablement

Your team runs it when I leave. Documentation, ownership and the operating cadence - a GTM backlog, sprint reviews, a definition of done.

What counts as a signal

Anything observable that raises the probability an account is in motion. Dozens are trackable. These are the 15 I rate highest for B2B SaaS and fintech - and the value is in which combinations predict your revenue, weighted by your deal history.

Lead changed company

A user or champion from a past deal lands somewhere new - with budget and a mandate to change things.

Lead promoted to new role

Yesterday's end user becomes today's budget holder, and your CRM already knows what they care about.

Contact departure

A champion leaves a customer account: churn risk on one side, a warm door wherever they land.

Company hired key roles

A new exec in the function you sell into. New leaders audit the stack they inherit and spend to make their mark.

Careers page updated

Hiring for the roles your product serves means budget and an initiative - visible months before any form fill.

Closed funding round

Budget arrives - and so does the competition. Pair it with a second signal and reference the plan.

Acquisition or merger

Two stacks become one, and duplicated tool categories get re-decided in the consolidation.

New market entry

Expansion multiplies operational pain - compliance, localisation, multi-entity, multi-currency.

Technographic profile change

A complementary tool added, or a competitor's tool removed. Either way, the stack is in motion.

Leadership or board change

A new board member or C-level hire signals a shift in strategy - and a window before priorities harden.

Headcount shift by department

Growth or contraction in your buyer's function tells you where the pressure is building.

Pricing page updated

A company changing how it charges is repositioning. Strategy is moving inside that business.

Regulatory approval secured

Compliance milestones create budget with a date attached - and open the markets that were waiting on them.

Partnership announced

Alliances reveal where a company is heading and which ecosystems it's betting on.

Lead engaging with industry content

A pattern of engagement from an ICP-fit person shows you the topic they're building a case around.

Alongside these market signals sits the layer only you can see: product usage, website and pricing-page intent, CRM history. That first-party layer carries the most predictive power, and it's where I start every build.

Why now

70%

of the purchasing journey is already complete before most B2B buyers make first contact. Signals are how you see the deal before the form fill.

Salespanel, 2025
79%

of B2B leads never convert to sales - and poor scoring is one of the most cited causes. The cost of unweighted signals is your team's focus.

Cognism, 2026

Teams stacking 4 or more signals per play report roughly double the reply rates of single-signal outbound.

Unify GTM research, 2025

Who this is for

CEOs, CROs and CMOs at B2B SaaS and fintech companies - typically Series A to C - with signals worth capturing: a product generating usage data, meaningful website traffic, a CRM with deal history. If you're pre-product or pre-traffic, start with the growth audit instead. It will tell you what to build first.

I've been building GTM systems on top of language models since 2020, across 20 years in B2B marketing. This is engineering-led advisory: I design the system with you, build it with your team, and leave you owning it.

Design the system first

A 30-minute conversation will tell you whether your signal problem is capture, scoring or acting - and whether you need a build, an audit, or neither.

Frequently asked questions

What is signal-based GTM?

Running your go-to-market motion on evidence that accounts are in market - product usage, website behaviour, champion job changes, hiring patterns - rather than on static lists and volume outbound. In practice it means infrastructure: capturing the signals your business generates, weighting them against your own deal history, and wiring qualifying combinations to specific plays.

We already have intent data. Why isn't it producing pipeline?

Usually for 3 reasons. The data is third-party, so your competitors are buying the same feeds. The scoring is a vendor model trained on other companies' pipelines, so reps have no reason to trust it. Then there's the follow-up: a score moves in the CRM and the next step has no owner. The compounding results come from fixing all 3 as one system.

Do we need Clay, Common Room or another signals tool?

Sometimes. They're strong capture layers, and I use several of them in client builds. The tool is the least important decision, though. Buying one before designing the system is how teams end up with an expensive spreadsheet that has no owner. Design first: which signals predict your revenue, weighted how, triggering what plays, owned by whom. Then choose tooling to fit the design - sometimes that's Clay, and sometimes your warehouse and CRM can do the same job without a new subscription.

Should we build this or buy it?

The capture layer is often worth buying - visitor identification, enrichment and job-change tracking are commodities. The scoring layer is worth owning, because its value comes from training on your closed-won and closed-lost data, which no vendor has. The acting layer lives in the tools you already run - CRM and sequencing. Buy the plumbing, own the scoring, and run the plays through tools you already have.

How is this different from buying a lead-scoring feature?

A lead-scoring feature is a model someone else trained, applied to whatever data happens to reach your CRM. Signal infrastructure covers the whole chain: capturing first-party signals that never reach your CRM today, weighting them on your own revenue history, and mapping score changes to owned plays with deadlines. That chain produces pipeline you can trace back to specific signals - and retrain as your motion changes.

Where do we start?

With an inventory. A signal audit maps what your product, website and CRM already emit against what you're paying vendors to approximate, and identifies the 2 or 3 signal combinations most likely to predict your revenue. It also tells you whether you need a build at all - some teams just need their existing data wired together and one play given an owner.