The go-to-market skills worth practicing when AI writes the outreach

Among his rules for hiring a VP of Sales now, Jason Lemkin, SaaStr’s founder, asks for leaders who are data-native, and puts the forecasting point conditionally: “They understand that AI forecasting can actually be accurate — if the inputs are right.” (Jason Lemkin, SaaStr, opens the source) The condition is the skill. Our reading: an AI forecast is only as honest as the stage, close date and buyer evidence behind each deal. In an interview, weight pipeline by how it has actually converted, and say early when the number falls short.

Asked what he looks for in a GTM engineer, Intercom’s Alexander DeMoulin starts with selling experience: “I want someone who has sat in the sales seat before” (Alexander DeMoulin, Clay blog, opens the source). He was describing one role at one company. Our inference for any GTM interview: AI can score and enrich every lead, but deciding which accounts deserve a person’s time takes knowing which deals actually close, at what size and how fast, and saying so when the busiest segment isn’t the best one.

Databricks’ CRO Ron Gabrisko argues that without understanding the product and how it works, “it’s hard to teach your customers how to get value from it.” (Ron Gabrisko, SaaStr, opens the source) That was his case for technical sellers. Ours is narrower: an AI brief can describe an account, but only discovery establishes who decides, what triggers the purchase and what it is worth. A strong answer checks the AI’s guesses against the buyer’s own words.

Clay’s first GTM engineer, Yash Tekriwal, probes a candidate’s AI fluency by asking, “What’s an example of something that AI is very good at doing, and something that it’s not very good at doing?” (Yash Tekriwal, Growth Unhinged (Kyle Poyar), opens the source) Owner’s CRO Kyle Norton describes a weekly QA meeting on his team’s AI tools: “We go through what needed to be escalated to a person and why.” (Kyle Norton, SaaStr, opens the source) Our reading: knowing where AI outreach fails is the skill. Replies are cheap to generate; meetings, pipeline and a sending reputation the whole team relies on are not.

Kyle Poyar, who writes the Growth Unhinged newsletter, describes the shift in software pricing this way: “We’re moving away from charging for access to software and toward a model of charging for the work delivered” (Kyle Poyar, Growth Unhinged, opens the source). That is a market observation, not hiring advice. Our inference: when AI does the work, seat counts can shrink as the product succeeds, so sellers and deal desks are judged on whether a price and a discount still hold at renewal, not only on closing this quarter.