Jai Kharbanda

A 30-day sales-coaching loop that forces a verdict

GTM AI & teaching 30 July 2026

Context. My sales coaching system was good at producing analysis, but didn’t produce any action, recommendation, or at the very least, help to triage the opportunity. A polished call note, for example, could still let weak opportunities survive if they didn’t force a verdict, or at least encourage a deeper review.

What I did. Rebuilt the post-call and weekly review flow around a 30-day control loop. Every commercial call now has to classify the account as Advance, Test, Nurture or Kill, using all the evidence we have for the buyer; pain, their pain, along with a score that uses MEDDPICC as the foundation. The Friday review then rolls the week into seven sections: the board, what to repeat, what to change, the pattern watchlist, next training focus, and what leadership can act on.

What happened. The shipped artefact encouraged new operating habits, encoded in a live sales-coach skill, key templates and an automation registry. That means the agent can no longer hand me a clever summary when the real answer is somewhere buried inside all the context.

What I would reuse. Force the machine to choose a small set of business actions. Analysis is useful only when it forces progress.

Account names, call details and internal CRM data removed. Numbers are process numbers only.

All case studies