GTM Engineering - More than hype?

GTM engineering is having a moment, which is usually a good reason to be suspicious of it. So the title here is a real question rather than a rhetorical one.

Tom runs go-to-market engineering at Conveo. The team is him and a set of agents. Dieter asked him what the job actually consists of, what it has replaced, and what it takes to get a commercial team to trust any of it.

What they cover

• “SaaS is dead,” and what sits in the gap. Tom picked the phrase up during three months at Anthropic. The argument goes: SaaS sells you workflows, agents sell you outcomes, and what is left in between is a thin layer of coordination. That layer used to live in SOPs and in whichever tools each person had strung together. Now it is software, and at Conveo it is one internal app called Super G.
• What the app is actually for. From first principles the GTM job is to build a relationship with everyone you meet, and a relationship is made of remembered touchpoints. So Super G keeps all of them, from first contact to a customer who recommends you to other people. Tom walks through the personalisation arms race that led here, from mail merges that knew your first name to automated notes about the post you just published, and argues that remembering is the part that was always doing the work.
• Why they build rather than buy. HubSpot has to serve hundreds of thousands of customers, so it makes trade-offs that do not fit any one of them exactly. Building it yourself means one high performer can work the way they want to, and the next one can work differently.
• Adoption is a people problem before it is a technical one. Making feedback almost too easy to give, then going and asking for more of it. Giving the agents distinct personalities, and having them open a conversation with something the person would actually answer. There is a good stretch on one self-described holdout who wanted nothing to do with any of it and came around.
• Where the time went. Someone on the sales team was spending around two hours a day inside prospecting tools: building a list, filtering it, loading it into the next tool, keeping it in step with their own outreach. That is automated now and coordinated across the company, inbound routes to the right person by vertical, and a next-action column makes the choice when there is too much that could be done. It compounds outside sales too, since a research impact lead can open a call already knowing every touchpoint that came before.
• Why humans stay accountable. From a manifesto Tom wrote: the AI-native company that wins is the one that maximises its human potential. You cannot hold an agent accountable, because the social, financial and legal contracts we run on only bind people. He takes that through to who you hire, why system builders matter more than one-off operators, and what he is coaching the people who feel threatened by it to do instead.