Where AI belongs in your go-to-market, and where it does not

The winners are not the companies using the most AI. They are the ones who worked out which three jobs to give it and left the rest alone.

Nearly every company I walk into is doing one of two things with AI. Either nothing, and quietly panicking. Or everything, badly, in nineteen tools nobody has audited.

The useful frame is not "where can we use AI". It is "which jobs in our go-to-market are high volume, low judgement, and currently done by somebody expensive". That is a short list and it is where the money is. Research and enrichment before outreach. First-pass creative variants for paid media. Summarising calls into something a human will actually read. Routing and triage.

Then there is the list where it does not belong, and this is the part that gets skipped. Anything where being wrong is expensive and hard to detect. Positioning. Pricing. Who you say no to. The final version of anything a customer will judge you on. Every relationship that is worth having.

The failure mode I see most is companies automating the top of the funnel until the volume looks impressive and the quality of every downstream conversation quietly collapses. You cannot see it in the dashboard for two quarters. By then the list is burnt.

The operating model matters more than the tooling. Who owns the output. What gets reviewed by a human and what does not. What happens when it is wrong. Companies that answer those three questions get compounding value. Companies that just buy seats get a mess with a subscription.

← All writing