Electrical wholesaler: cable, lighting, distribution boards, EV charging and fixings for electrical contractors and maintenance teams across Staffordshire and South Cheshire. 96 staff.
Verdict: 6 salespeople compared (2 left out: new starters or too few quotations). Win rates run 26% to 40%; median of everyone compared 34%. The biggest gap is Rep F, win rate: 26% against 36% for the others, worth £34,322 of gross profit a year if it closed fully, before the chance it works. Smaller gaps that could be chance (2 rate gaps smaller than two standard errors) aren’t valued. The 5 development actions planned are expected to add £21,963 a year (counting 50% of each gap closed, at the chance each works), for £12,980 in the first year (£1,050 of it cash; selling time is costed at the gross profit it would have earned): a first-year net of £3,492 (-£4,744 if it works half as well). Differences can come from territory, customers, product mix and lead mix as much as skill: go through each person’s own figures with them before planning anything. This is for development, not discipline.
Each step is set against the median of the other five people compared, never the best performer. Win rates run 26% to 40%; median of everyone compared 34%. Rep F wins 26% of quotations at the lowest margin and the highest discount, which points to selling on price; the Discount Leakage Audit goes deeper on that. Rep D quotes 57% of enquiries against 81% for the others but wins at least as often once quoted; the unquoted enquiries are probably the weaker ones, so they’re counted at half Rep D’s win rate. Rep A’s quotation rate and Rep B’s win rate sit a little below the others but within normal variation: not valued.
| Salesperson | Quotation rate | Win rate | Average order | Margin | Gross profit a year | Weakest step |
|---|---|---|---|---|---|---|
| Rep A 84 months in role · 70% of enquiries from existing customers | 77% | 40% | £8,458 | 22.5% | £182,700 | Margin likely mix |
| Rep B 60 months in role · 65% of enquiries from existing customers | 82% | 32% | £7,297 | 24.0% | £129,600 | Average order |
| Rep C 36 months in role · 60% of enquiries from existing customers | 73% | 38% | £6,917 | 25.5% | £126,990 | Average order likely mix |
| Rep D 48 months in role · 68% of enquiries from existing customers | 57% | 36% | £7,377 | 23.0% | £103,500 | Quotation rate |
| Rep E 30 months in role · 2 months absent · 58% of enquiries from existing customers | 81% | 30% | £6,500 | 24.5% | £133,770 | Win rate |
| Rep F 72 months in role · 72% of enquiries from existing customers | 83% | 26% | £8,269 | 21.0% | £90,300 | Win rate |
| Rep G 4 months in role · 30% of enquiries from existing customers | 67% | 22% | £5,778 | 22.0% | £34,320 | Left out: new starter |
| Rep H 18 months in role · 85% of enquiries from existing customers | 53% | 44% | £8,714 | 27.0% | £16,470 | Left out: small sample |
Rep E’s figures are scaled to the ten months worked (two months’ shared parental leave). Rep G (four months in role) and Rep H (16 quotations, half their time on branch management) are left out. Rep A and Rep C have their weakest step in margin or average order but win more often than the others: that is more likely product or customer mix, or pricing to win, than a skill gap.
| Action | Expected gross profit a year | First-year cost | First-year net | Half as well |
|---|---|---|---|---|
| Joint visits with the sales manager, one day a month, and a review of lost quotations Rep F · win rate · tested elsewhere, 50% chance | £8,581 | £3,685 | £2,750 | −£467 |
| Weekly one-to-one on the enquiry log: every enquiry quoted within 48 hours or closed with a reason Rep D · quotation rate · worked here before, 60% chance | £6,676 | £2,466 | £2,541 | £37 |
| Quotations below 20% margin checked by the sales manager before they go Rep A · margin · best practice, 30% chance | £1,827 | £1,134 | £237 | −£448 |
| Manufacturer training on lighting controls and EV charging, to quote the whole job Whole team · average order · best practice, 30% chance | £4,188 | £3,875 | −£734 | −£2,305 |
| Enquiry follow-up call within a day of a site visit Rep C · quotation rate · hunch, 20% chance | £691 | £1,820 | −£1,302 | −£1,561 |
Only the enquiry-log one-to-ones for Rep D and the joint visits with Rep F clearly pay. Rep A’s margin check is aimed at what looks like mix, and could cost some of the wins that make Rep A the team’s best closer: talk it through first. Rep C’s follow-up calls (a hunch) and the team’s product training cost more than they bring in the first year, mostly in selling time given up; run the training only if the manufacturer will do it outside selling hours, or for other reasons. Rep E’s win-rate gap has nothing planned yet. The plan needs 134 of the sales manager’s 160 available hours.
Who to learn from: Quotation rate: Rep F 83%; win rate: Rep A 40%; average order: Rep A £8,458; margin: Rep C 25.5%. Pair people with whoever is strongest at their weakest step for a day’s shadowing, and write down what that person does; the strongest is someone to learn from, not a target.
Each person’s enquiries, quotations, wins, sales and margin come from the sales records for 12 months, scaled to the months each actually worked; anyone with fewer than 50 quotations or under 6 months in role is left out. Each step is set against the median of everyone else compared, and the gap valued on a year’s gross profit; rate gaps smaller than two standard errors are treated as chance and not valued, and extra quotations are assumed to win at 50% of the person’s own rate (assumed). Development is assumed to close 50% of a gap (assumed), at the chance each action works (assumed: 60% worked here before, 50% tested elsewhere, 30% best practice, 20% a hunch), building up over 6 months; selling time given up is costed at each person’s gross profit per selling hour. Steps interact (quoting more can lower the win rate; a higher margin can lose orders), and territory, customers, product mix and lead mix differ, so a gap is a reason for a conversation, not a measure of ability.
Share each person’s own figures with them before anything is decided, and ask what the figures miss: territory, the customers they inherited, enquiries passed on late, product mix and copper prices. Keep the file secure and use it for development only (UK GDPR, with a documented legitimate interests assessment and the team told in a privacy notice). This isn’t legal advice: follow the ACAS Code of Practice and take HR advice before any figure is used in a formal process.
At handover, Corvanne's sales manager went through each person's figures with them and started the development actions. The agent was then set up in Corvanne's own AI assistant account, with read-only access to the sales system. After each quarter it reruns the review on the latest 12 months in a new dated copy, with the months absent confirmed by the sales manager, runs the checks and sends the sales manager a short note of what changed for each person since the last review; on the first Monday of the other months it checks the connection still works. It uses the rules on Inputs: 50 quotations and 6 months in role for a fair comparison, and a £10,000 rise in a person's gap before it flags one. It never read absence or HR records and never contacted a salesperson. Nothing left Corvanne, and no decision changed, without one of its own people approving it. These events continue the fictional example.
| Date | What the agent found | Flag? | What people decided |
|---|---|---|---|
| 2 Nov 2026 | Test rerun on the latest quarter's figures, the 12 months to 30 September, with the first monthly connection check: every figure matched the delivered review and all checks passed. | Note only | Nothing needed. Kept for the 30-day check-in. |
| 4 Jan 2027 | First quarterly rerun. Rep D's quotation rate rose from 57% to 66%, and their weakest step changed to the win rate. | Flag | The sales manager went through it with Rep D: the weekly enquiry-log one-to-ones were working. They kept them and added a review of lost quotations. |
| 1 Feb 2027 | A system update had renamed the quotations field, so a rerun would have shown no quotations for anyone. The agent sent a data problem note and no figures. | Flag | The system administrator mapped the renamed field; the test export passed two days later. |
| 5 Apr 2027 | Rep G passed 6 months in role and 50 quotations, so was compared for the first time. | Flag | The sales manager went through Rep G's figures with them before anyone else saw them, and started monthly joint visits. |
| 5 Jul 2027 | Rep F's gap at the win rate rose by £14,600 a year, to £48,900, more than the £10,000 set on Inputs: two framework tenders lost in May. | Flag | Both were lost on price to a national distributor, not on the sales process. No new development action; the joint visits continue. |
| 4 Oct 2027 | Nothing crossed a line, and sales were within 10% of the accounts. | Note only | Nothing needed. The six-monthly review with each person went ahead as planned. |
At the 30-day check-in on 16 November, Geoff and the sales manager read the test rerun on the 12 months to 30 September together and confirmed it matched the review. No rule needed changing. Geoff's upkeep ended there: the first real quarterly rerun, in January, and the agent and its log were Corvanne's.
What the agent added. The method is unchanged: same comparison, same fairness rules, same development plan. The agent made sure the figures were rerun every quarter and that a change for any one person reached the sales manager, who went through it with them. The benefit here is development that follows the evidence; no £ figure is claimed for it.
Fictional and labelled: these events continue the worked example. The same 6 events are the sample rows in the Agent log of the worked-example spreadsheet. The agent is part of the guided and done-for-you routes; the self-implement kit does not include it.