Signal-Based Outbound: Prospecting After Spray-and-Pray Died

Let me start from the beginning, because the whole argument depends on the sequence of events. Through automation, in the late 2010s and early 2020s, basically every written channel drastically lost its potency. Reply rates collapsed, because people now assume any cold message is spam. The sheer volume keeps climbing. Worldwide, people sent and received about 281 billion emails a day in 2018; Statista projects that reaches 392.5 billion in 2026 and over 424 billion by 2028. More mail, more automation, more reluctance.

This is for founders and sales leaders who can feel that "just send more, but personalised" has stopped working, and who want a model that holds up in 2026. It draws on my own research into how outbound is actually run in Europe, not on a vendor's playbook.

1What ChatGPT and Clay Actually Changed

Two things shifted the ground. First, with the advent of ChatGPT, we could finally go beyond placeholder-based personalization, the "Hi {first_name}, I saw {company} is in {industry}" template, and write full prompt-based messages that, most of the time, didn't sound robotic. Second, with Clay and tools like it, we could combine many sources of information to do two things: build lists at scale that are hyper-personalised and hyper-relevant, and feed signals and research into a prompt that spits out a tailored message.

That's genuine progress. But it created an arms race, and the arms race is the whole problem. When everyone can generate a relevant-sounding message at volume, a relevant-sounding message at volume stops being an advantage.

2Relevance Beats Personalization: The Window of Opportunity

Here's the realization that signal-based outbound is built on: being personalised doesn't cut it. Even if you've done your research, the message has to be relevant to the buyer's context right now. There's a window of opportunity, and it's usually signalled to the outside world by a handful of symptoms that, if you understand your persona really well, you can capture with modern tools.

3The Go-To-Market Alpha

At best, you combine these signals into something Clay has called a go-to-market alpha: a unique signal, or a unique combination of symptoms, that truly speaks to the value proposition you're selling and therefore signals very high relevance for a specific customer.

And to be honest, it works. If you're worried about a specific problem right now and a message lands speaking exactly to that problem, you'll pay attention, even if it's a mass email, even if it's not perfectly researched. That's just how we're wired.

4The Catch: Signals Are Commoditized

Now here's the thing. Signals and symptoms are, a lot of the time, commoditized. What you think gives you a unique insight is probably being used by every other company now running signal-based outreach. Even if you're relevant, you get crowded out and end up as noise anyway, because there's simply too much messaging out there and too many people holding the same tools.

And it gets worse when you look at who designs these campaigns. The people building signal-based outreach are go-to-market engineers, and in my own analysis of European go-to-market engineers on LinkedIn, the majority come from a marketing background and hold no negative inclination toward mass mailing. If the only tool you have is a hammer, everything looks like a nail. They're trained in one-to-many, one-to-volume outreach, and they're most comfortable with volume, because if you think like a data engineer, you want a lot of volume to figure out what's working and what isn't.

That's exactly where signal-based or symptom-based outreach falls down. You end up directing a fire hose at full power toward a very fragile flower, which is the lead. And it's not one fire hose. It's multiple go-to-market engineers doing the same thing at the same time, from multiple angles, all aiming at the same flower. You don't water it. You drown it.

I'm not saying signal-based outbound doesn't work. Relevance will prevail. But every channel that can be automated is losing its potency, because the volume wizards are spamming it and need volume to make it work.

5The Fix: Innovative Pipeline, Conservative Channels, Human Outreach

What I'm arguing for is to use go-to-market engineering but to constrain it: a conservative approach to volume and an innovative way of thinking about channels. In practice, that's three moves.

Build a richer go-to-market alpha, including first-party signals

By all means, build your go-to-market alpha. But don't think of it only as external information, like job postings on the customer's site. Also put out lead magnets that signal pain awareness, think webinars, whitepaper downloads, or any first-party intent capture, that in combination with external signals make your go-to-market alpha even more unique and more accurate. The combination of someone's first-party behaviour and their external symptoms is far harder for a competitor to replicate than a job posting everyone can scrape.

Don't automate the volume just because you can

Once you have that alpha, the temptation, especially if the person owning the process has a go-to-market engineering background, is to automate the volume and blast out messaging. Resist it. Instead, think of innovative ways, ideally ways that can't be fully automated, to reach out.

Use channels that resist automation

The channel that stands out here is cold calling. It's been presumed dead for a decade and has never actually been replaced. With most of my clients it still books 80 to 90% of the meetings, and doing it well is a true skill. Physical mailing and partnerships sit in the same non-automatable category. You can still use the occasional LinkedIn message or email, but only with a human in the loop.

6Cold Calling Needs Volume. Spam Doesn't.

One nuance that trips people up: I'm against volume in spam, not against volume everywhere. In per-person outbound calling, you actually do need volume, because without it a whole host of bad habits creep in and your conversion rates drop, not only in absolute terms but on a relative, per-call basis too. Fewer calls and more research sounds disciplined, but below a certain volume the skill itself decays. So the prescription is precise: low volume of automated written spam, high volume of skilled human calling.

That's why every written touch goes through a human-in-the-loop approach: the tool suggests the message, a person edits it, and a person sends it. The cost of the occasional bad automated send is not a slightly lower reply rate. It's the permanent loss of a hard-won, high-relevance lead.

The Whole Argument in One Line

An innovative approach to building, segmenting, and enriching your pipeline. A conservative approach to which channels you use. And a deliberately human approach to the outreach itself. That combination is the thing that still wins, precisely because it's the thing the volume wizards can't copy at scale.

Win Where the Volume Wizards Can't Follow

The teams that win in 2026 aren't the ones sending the most relevant-sounding messages at scale, because that's the exact game everyone is now losing. They build a sharper go-to-market alpha, stay disciplined about channels, and keep the outreach human, with cold calling at the center. That's a coaching and process problem as much as a tooling one, and it's what my prospecting training is built around.