Blog & News

chevron_right

GTM Engineering

chevron_right

September 25, 2026

 ·  

4

min read

What to automate in GTM, and what to keep human. Where I draw the line.

By

Jimmy

McLellan

Co-founder · FCMO, Full-stack Marketer

On This Page:

    TL;DR

    Automate the prep: build target lists, enrich every account, and track signals like a new hire, a promotion, fresh funding or someone posting about the exact problem, with instant alerts to the owner. Being first to the moment something changes is often the whole game. Refuse to automate the message to a real person, the reply to a warm lead, and the judgement. The rule: automation should end with a well-briefed human, not a sent message.

    Where's the line between automate and keep human?

    Automation should end with a well-briefed human, not a sent message. That's the whole rule, and it's the cleanest test I know. If your system's last step is a person picking up a fully prepared, perfectly timed opportunity, you've built it right. If its last step is firing off a message to a real prospect on your behalf, you've drawn the line in the wrong place.

    Everything below is really just that rule applied. Automate the prep, keep the connection human, and make the handoff between them as rich as you can.

    What should you automate first?

    The account-based prep. Building target lists, enriching every account, and tracking signals. This is the repetitive, high-volume, no-judgement work that eats a person's week and that machines are genuinely better at anyway.

    Start with the list. Define who you're going after and build it properly, then enrich every account so you're never staring at a blank row: firmographics, the buying group, the context, the recent moves. A human researching one account well takes half an hour. A good enrichment setup does it for a thousand accounts overnight and never gets bored or sloppy on account four hundred. If you want the specific tooling that does this without overspending, I've laid out the modern GTM stack for a lean team separately.

    Why do signals matter so much?

    Because being first to reach out the moment something changes is often the whole game. B2B timing beats B2B cleverness most weeks. The best-written cold email into a company that isn't thinking about the problem loses to a plain, human note that lands the day the problem became urgent.

    So you set signals and have them notify the owner instantly. A new senior hire in the function you sell to. A promotion that changes who holds the budget. Fresh funding that unlocks spend. Someone posting, in public, about the exact problem you solve. When one of those fires, the account owner should know within minutes, with the context already attached, so a person can reach out while it still matters.

    Timing is the edge. Automate the watching so a human can be first to the moment that counts.

    Note what's automated and what isn't there. The watching, the enrichment, the alert: all machine. The reach-out itself: human, every time. The system's job is to put a person in exactly the right moment with everything they need. It's not to do the reaching out for them.

    What should you refuse to automate?

    The message to a real person, the reply to a warm lead, and the judgement. Those three stay human, and I'd argue about it with anyone.

    The first message matters because it's the moment a stranger decides whether you're worth their attention, and an automated one almost always reads like exactly what it is. The reply to a warm lead matters more, because someone has raised their hand and the single worst thing you can do is answer a human moment with a machine. And the judgement, who's worth pursuing, what to say, when to walk away, is the actual skill; it's the bit you're paying a good operator for. Automate the judgement and you've automated away the reason the role exists.

    Haven't people taken this too far?

    Yes, and it's worth naming. People have pushed GTM engineering too far up the pipeline, automating their way into what is really an account manager's job. Fully automated nurture of a live relationship. Bots replying to warm leads. Sequences running the conversation after someone's already engaged. That's silly, and it costs you the exact thing that wins in B2B.

    Human connection is the point. People buy from people, especially on considered B2B purchases with real money and real risk attached. The further up the funnel you go, from cold prep towards a live human relationship, the less you should automate, until you're not automating at all. Automation earns its place doing the grunt work that gets a well-briefed human into the right room at the right time. After that, get out of the way. If you're wiring this into your CRM, the same discipline applies to CRM automation: automate the plumbing, never the relationship.

    If you want a hand drawing this line in your own pipeline, so the machine does the prep and your people do the parts that matter, book a free Growth Audit. Our GTM engineering page shows how we build to that rule.

    Frequently asked questions

    What are the best GTM signals to track?
    remove
    add

    Anything that changes urgency or who holds the budget. A new senior hire in the function you sell to, a promotion, fresh funding, or someone posting publicly about the exact problem you solve. Set these to alert the account owner instantly with context attached, so a person can reach out while the moment is still live.

    Can't I automate the first cold message if it's personalised?
    remove
    add

    You can, and it almost always reads like what it is. The first message is where a stranger decides if you're worth their attention, so it's the wrong place to save time. Automate the research and the timing that make the message land; let a human write and send the thing itself.

    Isn't fully automated nurture the whole promise of GTM engineering?
    remove
    add

    No, and that's where people take it too far, up the pipeline into what's really an account manager's job. Human connection is the point in B2B; people buy from people. The further you get from cold prep towards a live relationship, the less you should automate, until you're not automating at all.

    So what's the one rule to remember?
    remove
    add

    Automation should end with a well-briefed human, not a sent message. If the last step is a person picking up a prepared, well-timed opportunity, you've built it right. If the last step is a machine messaging a real prospect for you, redraw the line.

    By

    Jimmy

    McLellan

    Co-founder · FCMO, Full-stack Marketer

    TEch Stack

    Tools referenced in this article

    Related Expertise

    Specialties covered in this article

    More articles...

    Sort ascSort desc