AI killed the activity metric. Sales is about to get very honest.

Sales management ran for decades on a proxy: if outcomes are hard to attribute, measure activity. Dials, emails, touches, meetings booked. It worked because activity was expensive, so volume at least proved effort. The expense was real: reps in Salesforce's State of Sales survey of 5,500 sales professionals put 70% of their time on non-selling tasks.

AI made the volume nearly free. Account research that took an hour takes minutes with the file and the firmographics in the model. Outreach drafts personalize at scale. Call notes write themselves, the CRM updates itself, the follow-up ships before the prospect's elevator reaches the lobby. A rep with real AI fluency now produces a team's worth of motion alone, and it's mainstream: 81% of sales teams told Salesforce they're experimenting with AI or running it outright.

Which breaks the proxy. When everyone can do two hundred touches a day, two hundred touches stop meaning anything, and the spreadsheet finally has to measure what it always claimed to: pipeline created, deals advanced, revenue closed. The gap is already visible in the results, 83% of teams using AI grew revenue last year against 66% of teams without it, per the same Salesforce research. That honesty is the real disruption, and it cuts both ways.

Strip away the production and what remains is what great sellers always did: discovery that finds the problem behind the stated problem, deal strategy across a committee, trust built over time, and the triage call of which accounts deserve human hours at all. None of that automates, and all of it gets more valuable when the busywork stops swallowing the week.

The same split we've tracked in engineering and marketing applies: the rep becomes a director of machine output, briefing AI with real account context, editing drafts so they sound like a person who did the homework, and verifying before anything customer-facing ships. Buyers delete machine-sounding outreach faster than ever, and the flood of AI-written sequences is training them to spot it on sight, which makes the editing skill, not the generating skill, the differentiator.

Team math changes too, and the employment data has arrived. Stanford's Canaries in the Coal Mine study, built on ADP payroll records, found a 16% relative employment decline for workers aged 22 to 25 in the most AI-exposed occupations, a list that includes wholesale and manufacturing sales reps, while their colleagues aged 35 to 49 in the same exposure tiers grew over 8%. The study's sharpest finding actually favors sellers: employment fell where AI automates the work and held where it augments it. Closer to home, Emergence Capital's April 2025 survey of 560+ B2B software companies found 36% had cut SDR headcount in the past year, more than any other sales role. The call-blitz tier is shrinking the way entry-level roles are shrinking everywhere, and the path in for new sellers is arriving AI-fluent and learning judgment fast under a veteran.