Sales KPIs that matter: the six numbers a sales manager needs to see every week
Which sales KPIs actually steer a team: win rate per stage, sales cycle length, average deal size, pipeline velocity, contribution margin per salesperson and early churn rate — with formulas, a worked example, vanity metrics to drop and the legal framework for per-employee reporting in Germany.
The short answer
A sales manager needs six numbers every week: win rate per stage, sales cycle length, average deal size, qualified open opportunities, contribution margin per salesperson and early churn rate. The first four explain how fast the pipeline produces money. The last two tell you whether that money stays. I have seen sales organisations with forty KPIs in their weekly report that still could not say why the quarter went badly. More numbers almost always mean less steering, because nobody knows any more which one to move.
The test: what do I do differently on Monday if the number drops?
Every KPI in the report must answer one question: what action follows if it gets worse? If the win rate at the proposal stage drops, I look at proposals and price conversations. If the cycle gets longer, I check whether we are talking to the decision-maker too late. If no action can be traced back to a number, it leaves the weekly report. It may stay in the quarterly review, but not in the meeting where decisions are made.
The second test: can the salesperson influence the number? Market share, the industry cycle or delivery times belong in strategy, not in the assessment of individual people.
The six numbers and what they reveal
| KPI | Formula | What a drop usually means |
|---|---|---|
| Win rate per stage | won ÷ (won + lost) opportunities that reached this stage | A problem at exactly this stage, not “in sales in general” |
| Sales cycle length | average days from qualified first contact to signature | Decision-maker involved too late, unclear next step |
| Average deal size | order volume ÷ number of deals | Discounts, wrong target group, packages too small |
| Qualified open opportunities | opportunities that met the exit criterion of the first stage | Prospecting has fallen asleep; the effect shows with a delay |
| Contribution margin per salesperson | revenue minus variable costs and commission, per head | Revenue is being bought with price |
| Early churn rate | cancellations or withdrawals within the first 90 days ÷ deals | Things are being sold that do not fit |
The key words are per stage. An overall closing rate of 20 per cent tells you nothing. Whether the losses happen in the first meeting or after the proposal decides whether I have a qualification problem or a price problem. How to build stages with clean exit criteria is covered in the article on the pipeline review.
Pipeline velocity: four levers in one calculation
The first four numbers can be combined into one figure:
Pipeline velocity = opportunities × win rate × average deal size ÷ cycle length in days
An example with freely chosen values: 40 qualified opportunities, 25 per cent win rate, 6,000 euros average deal size, 60-day cycle.
40 × 0.25 × €6,000 ÷ 60 = €1,000 per day
| Lever improved by 10% | New value | Result per day |
|---|---|---|
| Opportunities | 44 | €1,100 |
| Win rate | 27.5% | €1,100 |
| Deal size | €6,600 | €1,100 |
| Cycle length | 54 days | €1,111 |
Mathematically, all four deliver almost the same. In practice they differ greatly in what they cost. Ten per cent more opportunities usually means ten per cent more prospecting effort. A cycle six days shorter often comes simply from having the decision-maker at the table in the first meeting instead of the third. The calculation therefore does not tell you where to start — it shows where the cheapest lever is.
Revenue is the wrong headline number
Many sales teams steer and pay by revenue. That rewards the salesperson who closes with a 15 per cent discount just as much as the one who holds the price, even though the first leaves the company considerably less. That is why contribution margin per salesperson belongs in the weekly overview and, where possible, in the commission logic.
The second addition is the early churn rate. If you only measure new deals on recurring contracts, you pay for sales that are gone again after three months. I have seen this in organisations that were proud of their closing figures while their customer base still did not grow. What happens in the first 90 days after the deal therefore belongs in the same report as the deal itself.
Vanity metrics: what you can drop
- Number of calls without a link to results. Useful as an activity target, worthless as a measure of success. A hundred calls to the wrong target group are just busyness.
- Total pipeline value. A sum that counts every untouched old opportunity grows by itself and says nothing about next month.
- Meetings without a qualification criterion. A coffee with an acquaintance is not a first meeting.
- Proposals sent. Only counts if the customer confirms receipt and a next step is agreed.
Counting activities is not wrong; it is just a different level. How outcome and activity targets fit together is explained in the article on sales targets.
Per-employee reporting: the legal side
As soon as the CRM automatically evaluates KPIs per salesperson, it is a technical system capable of monitoring performance or behaviour. If there is a works council, it has a right of co-determination under § 87(1) No. 6 BetrVG (German Works Constitution Act). This applies regardless of whether the manager intends to use the data for monitoring. Data protection law for processing employee data applies as well. My advice: write down the six KPIs, their purpose and who sees them beforehand, and agree them with the works council and data protection. That takes two weeks and saves you an argument over every single report later.
In the conversation with the team, one sentence that makes the purpose clear helps:
“These numbers are there so we can see where we lose deals — not to show who is at the bottom of the ranking.”
Action list
- Take your current weekly report and delete every KPI for which nobody can name an action if it drops.
- Report the win rate per pipeline stage, not as a total.
- Calculate pipeline velocity once with your own values and estimate, for each of the four levers, what a ten per cent improvement would cost.
- Put contribution margin per salesperson next to revenue and make discounts visible.
- For recurring contracts, add the early churn rate to the same report as the deals.
- Before any per-employee reporting, involve the works council and data protection and fix the purpose and the audience in writing.
FAQ
Which KPIs matter most in sales?
Six are enough for weekly steering: win rate per pipeline stage, average sales cycle length, average deal size, number of qualified open opportunities, contribution margin per salesperson and the share of customers who cancel or withdraw within the first 90 days. Everything else is root-cause analysis once one of these six numbers moves.
How do you calculate pipeline velocity?
Number of qualified open opportunities times win rate times average deal size, divided by the average sales cycle length in days. The result is the revenue or contribution margin the pipeline generates per day. Its value lies less in the absolute figure than in comparing which of the four levers is easiest to move.
Can sales KPIs be evaluated per employee?
Yes, but with conditions. As soon as a CRM or other software can evaluate the behaviour or performance of individual employees, an existing works council has a right of co-determination under § 87(1) No. 6 BetrVG (German Works Constitution Act). Data protection law applies as well. Clarify this before introducing personal reporting, with a lawyer if in doubt.