A sales performance flywheel has five parts: a standing question set built with your best manager, a rule that the debrief happens inside the recall window, the same set for every rep so the answers are comparable, a weekly coaching pass over those answers, and a quarterly review that sharpens the questions. Every part works manually, with a legal pad and a phone. The constraint is discipline, because someone has to ask, every day, forever. The automated version is Call-to-CRM: the rep calls June after a meeting, June asks the standing questions, and the answers land structured in the CRM you already run.
Below is the whole system. Build it yourself first. You will learn more from two months of running it by hand than from any vendor deck, including ours.
How do you build a sales performance flywheel?
Five parts, in order.
1. Write a standing question set with your best manager
Five questions, not fifteen.
A working set: next step and owner, decision process, risk, competition, amount and timing.
Do not write these alone. Sit down with the manager on your team whose deals are cleanest and write down what they actually ask on a ride-along. That is your set. It already works, it just is not written down anywhere.
Then resist adding to it. Every question you add costs you compliance on all the others.
2. Ask inside the recall window
The debrief happens on the drive out, not Friday afternoon.
Unrehearsed detail decays steeply, and most of the loss happens in the first hours. Murre and Dros replicated Ebbinghaus's forgetting curve in 2015 and landed within a few percentage points of his 1885 results (PLOS ONE). Those studies used nonsense syllables, not sales meetings, so take it as a mechanism rather than a number about your reps.
The mechanism is enough. A rep's read on a meeting is unrehearsed by definition. They did not write it down and they did not say it out loud. By evening, they are reconstructing.
3. Same set, every rep, every meeting
Comparability is the point.
You cannot coach across reps if every debrief is a different shape. When all of them answer the same five questions, patterns surface that no single deal review would show you. One rep never has a named decision process. Another names the competitor two meetings later than everyone else. Those are coachable. A pile of freeform notes is not.
4. Coach from the answers, weekly
Once the answers already exist, the 1:1 stops being discovery and starts being coaching.
That is where the returns show up. CSO Insights' 2017 Sales Enablement Optimization Study put the average win rate on forecast deals at 51.8%. Organizations that left coaching to managers at random landed at 43.6%. A formal coaching process landed at 58.8%, and a dynamic one at 66.1%, which the study frames as a 27.6% improvement in win rates (CSO Insights).
The coaching is the part that moves the number. The debrief is what makes the coaching possible.
5. Sharpen the questions quarterly
Once a quarter, read your losses against your answers.
When the same gap shows up in three lost deals, one question gets sharper. "Who else has to say yes" earns a permanent spot. A question that never produced a useful answer comes out. Then the new set is the standing set until the next quarter.
That is the loop closing, and it is the part most teams never reach, because they stop at part two.
What questions should be in a post-meeting debrief?
Next step and owner. Decision process. Risk. Competition. Amount and timing.
Start with five. Add a sixth only when a quarter of lost deals tells you to.
The temptation is to build a fifteen-field form because the CRM has fifteen fields. A rep in a parking lot will answer five questions honestly or fifteen questions carelessly.
When should the debrief happen?
Inside the recall window, right after the meeting, ideally from the car.
The walk from the conference room to the driver's seat is the last moment the rep's read is accurate rather than remembered. Anything captured after that is a reconstruction, and reconstructions are confident and wrong in the same places every time.
Why does the flywheel stop turning?
Almost never because the design was bad.
Most teams run it for about two weeks. Week one, everyone asks. Week two, a forecast call runs long, two reps are travelling, and Thursday quietly goes unasked. Nothing dramatic happens. The wheel just stops, and by the following Monday there is nothing to coach from, so the 1:1 goes back to being discovery.
That is not a discipline failure on anyone's part. It is what happens to any daily habit that depends on a person with a calendar. The design is sound. The asking is the part that runs out.
Which is worth knowing before you buy anything, because the software question is narrow: it is only about the asking.
What does the automated version look like?
That last part is what June does.
The rep calls June after a meeting. She asks your standing question set, holds a normal conversation for three to five minutes, and confirms what she heard. The answers land structured in your CRM, on the same fields, for every rep, every meeting. No app, no login, no typing.
Call-to-CRM, not voice-to-CRM. She asks the questions rather than waiting for dictation, and the question set is configured for your team rather than edited by each rep in a UI.
Build it manually first. If it works by hand, it will work automated, and you will already know exactly which five questions belong in it.
Frequently asked questions
What questions go in the debrief?
Next step and owner, decision process, risk, competition, amount and timing. Start with five and resist adding more.
When should the debrief happen?
Inside the recall window, right after the meeting, ideally from the car. By evening, the detail is reconstruction.
Do we need software for this?
No. You need discipline. Software is how the system survives the weeks discipline runs out.
How many questions is too many?
More than five, at the start. A rep answering from a parking lot will give you five honest answers or fifteen careless ones.
How long before it shows up in the numbers?
The coaching research is about formal, sustained processes rather than a fixed timeline, so treat the first quarter as building the habit and the second as reading the patterns.
Sources: Murre and Dros, PLOS ONE 2015; CSO Insights, 2017 Sales Enablement Optimization Study. Author: Shawn Johnson, Founder, CallJune.ai.