Marylou Tyler

External benchmarks can be tricky. What I like to do is first run an intra-rep analysis along segmentation lines (IB, OB, Referral, Self-Gen, Partner, etc.) to get a sense of variance. NOTE: I study top-to-middle-of-funnel, so make necessary adjustments for getting-to-closed analysis/diagnostics: By segment (or KPI filter..): INTERNAL BASELINES: If your top quartile reps are winning at 40% and bottom quartile at 15%, you know 40% is possible in your market with your product. COACHABLE GAPS: Study what the 40% reps do differently (qualification rigor, discovery depth, deal selection, etc.) SEGMENT SPECIFIC REALITIES: Data may naturally show which segments/deal sizes/use-cases/JTBD/personas have higher win rates for you. Then I'll look at variances and make assumptions that can be 'debated' with leadership team....

George C. I like your notion of auto-populating notes and activities using LLMs. Correlating pipeline velocity with adherence to workflow can now be [more easily] tracked and analysed using the unstructured data already in our CRMs. Next thing is the "naughty board" - a report that is generated with repeat offenders and correlates lack of compliance with intra- and inter-stage conversion rates + overall win / loss rates. Subliminaly tying tracking compliance to performance.

Max L. exactly - we're not tracking tasks like "proposal sent" for these 3-5 buyer intent signals. More like these behavioral characteristics:

I've only worked on POC-type demos, but here are the entrance criteria for those:

Exit criteria is extremely stringent and probably too long to list here, but the main categories are:

The best cold e-mail expert I know is Jason Bay - Outbound Squad. He has a lot of free material that can help you with your outbound campaigns. And, has an impressive track record. If you follow him, watch his free webinars, and put his teachings into action, consistently, with intent, you will be successful in your cold e-mail / voice / social campaigns.

Depends on the application but I do allow a weighting table that’s adjusted once we’ve got enough data for statistical relevancy. In competitive situations, losing control of the conversation is weighted higher (silence) than say latency. We also have seasonal indicators (bank holidays in Europe are especially tricky) so the algorithm doesn’t falsely report.

There are a couple schools of thought on sample sizes. If we have records, I like z-distribution to start focusing on 90% confidence. If we have too few records, then t-distribution with n=30 usually gives us ‘enough’ intel to decide whether a change is warranted. Key is what you’re willing to use as a confidence %.

CRM entries record intent. Buyer behavior reveals truth. And most clients still treat CRM entries vs. buyer behavior as the same signal. A diagnostic I like to run: :: Where does your CRM show progress while buyer signals show decline? :: Which behaviors (not fields) predict risk? :: What would change if every micro-signal triggered an action instead of relying on rep memory?

This prompts an important question: What early signals do sales teams notice before an opportunity goes dark? Common indicators I've seen often overlooked include: :: changes in communication frequency, :: shifts in decision-maker involvement, or :: sudden hesitations on previously agreed terms.

Churns & downsells aren't results — they're signals. Audit them like a black-box recorder, map where context drifted, and feed those insights directly into your renewal motions. That’s how you prevent or mitigate future churns/downsells.