Every sales organization tracks numbers. Most track too many of them, and a fair share track the ones that are easiest to pull rather than the ones that predict anything. A sales KPI is worth carrying only if a change in it would change what you do next week.
What follows is a working list of 20, with what each one measures, and where each tends to mislead. Take five or six of them, not twenty.
Metrics and KPIs are not the same thing
A metric is a count. Calls made, emails sent, meetings held. It tells you what happened without telling you whether it mattered.
A KPI is a metric you have decided to steer by, because it connects to an outcome the business cares about. Win rate, customer acquisition cost, and net revenue retention are KPIs when someone is accountable for moving them and a target exists. The same number on a dashboard nobody reviews is just a metric with better formatting.
The distinction matters for a practical reason. Activity metrics are the easiest things to measure and the easiest things to game, which is Goodhart’s law in its usual form: once a measure becomes a target, it stops being a good measure. Set a target of 60 calls a day and you’ll get 60 calls a day, many of them 20 seconds long. Activity numbers belong in coaching conversations, not in the compensation plan.
Getting this wrong is expensive in a specific way, because KPIs that feed sales management decisions usually also feed comp. A poorly chosen indicator doesn’t just mislead a forecast, it pays somebody for the wrong behavior. The same holds for measurement error: if the number is wrong, the payout is wrong, which is one of the more common incentive calculation mistakes.
Choosing which ones to track
Start from the business objective, not from the list. A company chasing recurring revenue needs net revenue retention, expansion rate, and churn. A company chasing new logos needs CAC, win rate, and pipeline coverage. Those are different dashboards, and running both because the software supports both is how teams end up with 30 numbers and no direction.
Bring finance into the selection. Sales and finance frequently define the same word differently, and a win rate calculated on qualified opportunities is a very different number from one calculated on every record in the CRM. Agree the definition and the data source before anyone builds a report. Pull two years of history from the CRM while you are at it, because a target set without a baseline is a guess.
Then cut the list to what a manager can actually hold in their head during a pipeline review. Five to seven is the usual working number. Review the set annually against what the business is now trying to do, and be willing to retire an indicator that has stopped moving.
Which ones matter also depends on what you sell. A distributor running rebate programs has to read revenue net of accrued rebates or the growth number is fiction. A team selling into a long, multi-stakeholder cycle gets more from pipeline velocity than from monthly bookings. The KPI set is part of the sales plan, not a separate exercise.
20 sales KPIs worth knowing
1. Revenue growth
Percentage change in revenue against the prior period. Useful only when you specify which revenue: gross bookings, net of rebates and returns, new versus expansion. A growth rate that mixes new business with renewals hides a shrinking new-logo motion behind a healthy base, which is the most common way this number lies.
2. Sales target attainment
Actual against quota, by rep and in aggregate. Read the distribution, not the average. A team at 100% where two reps carried the number and six missed has a different problem from a team at 90% where everyone landed close, and the average conceals both.
3. Customer acquisition cost
Total sales and marketing spend for a period divided by new customers acquired in that period. The argument is always about what goes in the numerator. Fully loaded CAC includes salaries, commissions, tooling, and overhead, and it’s the only version worth comparing to lifetime value. Ad spend alone produces a flattering number that means nothing.
4. Customer lifetime value
Expected gross margin from a customer over the whole relationship. Use margin, not revenue, or you’ll conclude that your least profitable segment is your best one. The ratio to CAC is what actually gets read: below 3 to 1 in a subscription business usually means acquisition is too expensive or retention is too weak.
5. Sales cycle length
Median days from first qualified contact to closed won. Use the median rather than the mean, because a handful of enterprise deals that took 14 months will drag the average somewhere no real deal lives. Segment it by deal size before drawing conclusions.
6. Lead conversion rate
The share of leads that become customers, best measured stage by stage rather than end to end. A single blended conversion rate cannot tell you whether the problem is lead quality, SDR qualification, or AE execution. Stage-level rates can.
7. Churn rate
The share of customers or revenue lost in a period. Track logo churn and revenue churn separately. Losing 10% of accounts that represent 2% of revenue is a segmentation issue; losing 2% of accounts that represent 10% of revenue is an emergency.
8. Monthly recurring revenue
Contracted recurring revenue normalized to a month. Its value comes from the breakdown: new, expansion, contraction, and churned MRR. A flat total can conceal heavy churn offset by heavy new business, which costs far more to sustain than the flat line suggests.
9. Average deal size
Revenue divided by closed deals. A rising average isn’t automatically good news. It often means the team has stopped working smaller accounts, which shows up two quarters later as a thin pipeline. Read it next to deal count.
10. Win rate
Closed won divided by total closed opportunities. Whether losses that were never really opportunities are counted changes this number enormously, so fix the qualification definition first. A win rate that climbs while pipeline creation falls means the team is only entering deals it already knows it will win.
11. Revenue per rep
Closed revenue divided by ramped reps. Exclude anyone still in ramp or the number punishes you for hiring. Compared against fully loaded cost per rep, it’s the clearest answer to whether the next headcount pays for itself.
12. Quote-to-close ratio
Quotes issued that turn into signed business. A low ratio points at pricing, at quoting too early, or at quoting to people without authority to buy. Track discount depth alongside it, since a high close rate bought with heavy discounting is a margin problem wearing a win rate costume.
13. Upsell and cross-sell rate
The share of existing accounts that bought more in the period. Expansion revenue carries almost no acquisition cost, which makes it the cheapest growth available and the first thing to check when new-logo numbers look expensive. Whether the comp plan pays for it is usually the reason it does or doesn’t happen.
14. Activity volume
Calls, emails, and meetings per rep. This is a diagnostic, not a target. When a rep misses quota, activity volume tells you whether the cause is too little pipeline or too little conversion, and those need opposite interventions. Attach a quota to it and it stops answering the question.
15. Pipeline velocity
Opportunity count multiplied by average deal value multiplied by win rate, divided by cycle length. The one composite worth computing, because it shows which of the four inputs to move for the largest effect on revenue per day. Often the answer is cycle length, which is also the one nobody works on.
16. Forecast accuracy
Forecast against actual, as a percentage variance, tracked per manager rather than per company. Persistent optimism and persistent sandbagging are both accuracy failures and both are visible within two quarters at the manager level. Bottom-up forecasts built from individual deals usually beat top-down targets, provided the stage definitions are enforced.
17. Net revenue retention
Revenue from the existing base at period end against that same cohort at period start, including expansion, contraction, and churn. Above 100% means the base grows without a single new customer. It’s the single most informative number in a subscription business and the hardest one to move.
18. Lead response time
Median minutes from inbound submission to first genuine contact attempt. Measure the median and the tail, because the average hides the leads that sat over a weekend. An automated acknowledgement isn’t a response and should not be counted as one.
19. Product performance
Revenue and margin by product line or SKU. Sales effort follows commission, so a line that reps ignore is often a line the comp plan underpays rather than one the market rejected. This is the standard reason product-level incentives exist.
20. Territory revenue growth
Growth by territory, read against territory potential rather than against other territories. A rep at 8% growth in a saturated region may be outperforming one at 20% in a region with four times the addressable accounts, and comparing the raw numbers rewards the wrong person.
Tracking them without a spreadsheet problem
Most KPI reporting fails on the data layer, not the visualization layer. Pipeline data sits in the CRM, revenue sits in the ERP, quota and payout sit in a workbook someone maintains by hand, and reconciling the three is a monthly manual job that produces a number nobody quite trusts.
Three categories of tool cover the ground. The CRM holds pipeline, activity, and conversion. BI tools such as Tableau or Power BI handle cross-system reporting once the sources are connected. Commission platforms, incentX among them, calculate attainment and payout from transaction data and keep an audit trail behind every figure.
Whatever the stack, apply two rules. One number, one owner, one definition, written down where the report lives. And every figure has to decompose back to the transactions that produced it. A dashboard nobody can reconcile to source becomes a dashboard nobody argues with, which sounds like agreement and is actually disengagement. That traceability is also what turns sales tracking into something operational rather than decorative, and it’s worth checking against your own forecast definitions before you build anything.
Where KPIs meet the comp plan
Not every KPI belongs in compensation. The ones that do share three properties: the rep can materially influence them, they can be calculated without judgment, and they survive being optimized against.
Revenue and margin attainment pass. Retention passes where the rep owns the account. Activity volume fails, because it’s trivially gamed. Forecast accuracy fails as a payout trigger for a different reason: paying on it teaches reps to sandbag, and you lose the forecast to protect the bonus.
Two or three components is the practical ceiling. Beyond that reps stop being able to work out what maximizes their own pay and revert to whatever they were doing before, which means the plan has cost money and changed nothing. Keep the earnings mechanics explainable in a single conversation, publish attainment against target on the same cadence you pay, and treat the plan as one of the revenue leadership decisions it actually is rather than an administrative afterthought. When the numbers are right, the same infrastructure supports commission calculation and reporting off one source.
Common ways this goes wrong
Too many indicators. Thirty numbers on a dashboard is the same as none, because nothing is prioritized. Cut to what changes behavior.
Definitions that drift. Two teams reporting different win rates from the same CRM is a definition problem, and it usually surfaces during a board meeting. Write the definition and the query next to the metric.
Lagging indicators only. Revenue tells you what already happened. Pipeline creation, conversion by stage, and cycle length tell you what is about to happen, which is the only kind of warning you can act on.
Averages without distributions. Nearly every KPI here means something different once you look at the spread. Team averages are how a two-rep dependency stays invisible until one of them resigns.
Targets set without a baseline. A number pulled from what the business needs rather than from what the team has historically produced does not motivate anyone. It just tells reps the plan is not serious.
Never retiring anything. Indicators accumulate. Review the set once a year and remove the ones that no longer inform a decision, the same way you would review territory design or demand planning assumptions.
The test for any indicator is the same one every time. If it moved five points, would anyone do anything differently? If not, it’s reporting, and it belongs in an appendix rather than on the wall. That test applies as much to financial forecasting inputs and revenue management reporting as it does to the sales dashboard, and to the incentive compensation measures built on top of them.
