Key Takeaways
- 1. The choice of measure is the most important plan design decision. It directly determines what behavior gets rewarded. Everything else (mix, curve, caps) is tuning. The measure is the signal.
- 2. Three measures is the practical maximum. Beyond three, reps cannot internalize priorities and will optimize for whichever is easiest. For most roles, one or two measures is ideal.
- 3. No measure should be weighted below 15% of target variable. Below that threshold, the payout impact is too small to influence behavior. You are adding operational complexity for zero behavioral return.
- 4. What you choose NOT to measure is as important as what you measure. The comp plan should not try to cover every business priority. Use management, process, and other incentive mechanisms for the rest.
If OTE determines how much you pay and mix determines how much is at risk, the measure determines what you are paying for. This is the design decision that most directly shapes behavior. Get the measure right and reps will naturally do the thing you need them to do. Get it wrong and they will rationally do the thing you are paying for, which is something else entirely.
Choosing the right measure requires answering three questions: What outcome does the business need most? Can the rep directly influence that outcome? And can we measure it accurately and frequently enough to calculate fair payouts? A measure that fails any one of these tests is a measure that will create problems.
The measure landscape
Revenue and bookings
The most common primary measure and the default for good reason. Revenue is unambiguous, directly tied to the P&L, and easy for reps to understand. "Close more deals, earn more money" is the clearest possible signal. The downside: revenue alone does not differentiate between a $100K deal at full margin and a $100K deal at 5% margin. If margin matters, revenue alone will not get you there.
New ARR / New Bookings
For SaaS companies, new ARR isolates growth from renewals. This prevents AEs from coasting on existing revenue and forces focus on new customer acquisition or expansion. The trade-off is that renewal revenue still matters for the business but is not incentivized through this measure, requiring a separate role or measure for retention.
Gross Margin
Margin-based comp works when reps have meaningful control over pricing and deal structure. In environments where pricing is fixed or approved centrally, paying on margin creates frustration without changing behavior. The rep cannot influence what they cannot control. Margin also requires clean cost data at the deal level, which many companies lack.
Activity metrics (meetings, pipeline, outbound)
Appropriate for SDR/BDR roles where the outcome (closed revenue) is owned by someone else. Activity measures should be paired with quality gates: not just "meetings booked" but "qualified meetings that the AE accepted." Without quality controls, activity measures produce volume without value.
Retention and expansion (NRR, churn rate)
The right measure for account management and customer success roles. Net Revenue Retention (NRR) captures both the downside (churn) and the upside (expansion) in a single metric. The challenge is measurement frequency: NRR is typically calculated annually or quarterly, which can delay the feedback loop for monthly-paid roles.
Revenue and bookings are lagging indicators: they measure what already happened. Pipeline, meetings, and activity are leading indicators: they predict what will happen. Most comp plans should pay on lagging indicators (outcomes) because they represent real business value. Leading indicators are better suited as gating measures (you must hit X pipeline before your revenue payout kicks in) or as primary measures only for roles that do not own the close (SDRs).
The 3-measure maximum
This is one of the strongest rules of thumb in compensation design, and I defend it vigorously. No plan should have more than three measures. Here is why.
A single-measure plan sends one clear signal: do this. The rep knows exactly what matters. A two-measure plan sends a primary and secondary signal: do this mainly, and also do that. Most reps can hold two priorities in their head. A three-measure plan is the edge: primary, secondary, and tertiary. The rep can usually manage if the weights clearly differentiate importance (60/25/15 or similar).
At four measures, the signals start to blur. At five, they are noise. At seven (yes, I have seen seven), the plan is not a motivation tool. It is a spreadsheet exercise that happens to determine pay. Reps default to the measure they understand best or the one that is easiest to influence, regardless of weight.
A company came to me with a plan that had five measures: revenue (40%), new logos (20%), margin (15%), customer satisfaction (15%), and pipeline generation (10%). I asked the VP of Sales to rank the measures by importance without looking at the weights. He got revenue right but swapped the order of everything else. If the VP who designed the plan cannot rank the priorities correctly, what chance does a first-year AE have? We cut to two measures: revenue (70%) and new logos (30%). Margin was addressed through pricing approval workflows. Satisfaction was moved to the CS team's plan. Pipeline was managed through activity expectations, not comp. The plan was clearer, cheaper to administer, and produced better outcomes.
Weight allocation: the 15% floor
Once you have chosen your measures, you need to allocate weights. The rules are straightforward:
Primary measure: 50-70%. This is the thing that matters most. It should be obvious to every rep which measure carries the most weight. If your weights are 40/35/25, no measure clearly dominates and reps cannot tell what to prioritize.
Secondary measure: 20-35%. This is the counterbalance. It ensures reps do not optimize for the primary measure at the expense of something else that matters. Revenue as primary, new logos as secondary is a common pattern: you want revenue, but not at the expense of customer concentration.
No measure below 15%. Below 15%, the dollar impact on variable pay is too small to change behavior. For an AE with $80K variable, a 10% weight means $8K at target. The incremental effort required to move from 90% to 110% on that measure might earn the rep an extra $1,600. That is not enough to drive behavioral change. If a measure is not worth 15%, it is not worth measuring in the comp plan.
A company paid on revenue (40%), margin (20%), new logos (20%), customer satisfaction (10%), and pipeline quality (10%). The satisfaction and pipeline measures each represented $8K at target. Reps rationally ignored both. The company spent considerable operational effort tracking, calculating, and reporting on two measures that influenced zero behavior. When they removed both and redistributed the weight to revenue (60%) and new logos (40%), the plan became simpler, cheaper to run, and produced identical customer outcomes because satisfaction was never going to be driven by a comp plan anyway.
The MBO trap
Management by Objectives (MBO) components appear in many plans as a catch-all for qualitative goals: "complete training," "maintain CRM hygiene," "demonstrate company values." They are almost always a mistake in a sales comp plan.
MBOs suffer from three problems. First, they are subjective. A manager's assessment of whether a rep "demonstrated leadership" is inherently inconsistent across managers and reps. Second, they are binary or near-binary in practice. Most managers give most reps 90-100% on MBOs because the political cost of a low score outweighs the financial savings. Third, they are impossible to connect to business outcomes. You cannot draw a line from "completed Q2 training modules" to revenue impact.
If you need reps to do something that cannot be measured quantitatively, use management accountability (it is your manager's job to ensure you complete training), performance reviews (CRM hygiene is a factor in promotion decisions), or gates (you must maintain 95% CRM compliance to be eligible for your accelerator). Do not use the comp plan as a behavioral swiss army knife.
What NOT to measure
Deciding what not to measure is often harder than choosing what to measure, because stakeholders always want their priority reflected in the comp plan. Here is the framework for saying no.
Do not measure things the rep cannot influence. If pricing is set centrally, do not pay on margin. If customer success owns onboarding, do not pay the AE on time-to-value. Paying someone on an outcome they cannot control is not motivation. It is a lottery.
Do not measure things that are expensive to track. If calculating a measure requires manual data collection, cross-system reconciliation, or subjective assessment, the operational cost may exceed the behavioral benefit. A clean, automated measure at 90% accuracy is better than a perfect measure that takes a week to calculate.
Do not measure things that should be table stakes. CRM hygiene, forecast accuracy, attending team meetings: these are job requirements, not bonus-worthy achievements. Including them in the comp plan sends the message that the company expects reps not to do these things unless paid extra. That is the wrong message.
Before accepting a new measure into the plan, ask three questions: Can I get clean data for this measure within 3 business days of period close? Can I calculate attainment for every rep without manual intervention? Can I explain the calculation in one sentence? If the answer to any of these is no, push back. A measure that breaks your operational process is a measure that will produce errors, delays, and disputes.
Are your measures aligned with your strategy?
Upload your comp plan and attainment data to see if your measures are actually driving the behavior you want. SalesComp Edge analyzes payout correlation and flags misalignment.
Try SalesComp Edge (free) Model your measures🤖 Try This Prompt
You are a sales compensation expert helping me choose the right measures for my comp plan. Here is my context: Business model: [Transactional / Land-and-expand / Enterprise / Channel] Sales cycle length: [Days/months] Current measures (if any): [List them with weights] Top 3 business priorities this year: [List them] Number of reps: [Count] Role type: [AE / SDR / AM / SE] Based on this context: 1. Recommend 1-3 measures with specific weight allocations 2. For each measure, explain: why it aligns with my business model, what behavior it will drive, and what risk it creates 3. List any current measures I should consider removing and why 4. Flag if any recommended measure would be operationally difficult to track 5. Suggest one gating measure (if appropriate) that should not carry its own weight but should serve as a qualifier
Need help aligning measures to strategy?
Book a 20-minute measure review. We will look at your current plan and recommend a simplified measure set aligned to your business priorities.
Book a measure reviewChapter Checkpoint
Test your understanding of measure selection.
Common Practitioner Questions
Only if they genuinely own the post-sale relationship. If you have a separate CS or AM team that manages renewals, the AE has limited influence over retention and should not be paid on it. If the AE owns the full lifecycle (common in smaller companies), a retention or NRR component can work as a secondary measure at 20-30% weight. The test: can the AE take specific actions that directly improve retention?
Use a gate when you want to ensure minimum performance on a metric without dedicating variable pay to it. Example: "You must maintain $200K in pipeline at all times to be eligible for your revenue accelerator." The pipeline is not directly compensated, but failing to maintain it blocks the upside. Gates are ideal for leading indicators, compliance requirements, and table-stakes behaviors that should happen regardless of comp.
Yes, and it is often the best design. A single-measure plan is the simplest, clearest, and easiest to administer. It works when your primary metric (usually revenue or new ARR) captures most of what you need. The risk is that reps optimize for that one metric at the expense of everything else. Mitigate this with gates, management accountability, or a well-designed accelerator that rewards quality over volume.
Show the math. Calculate the dollar impact of each measure at target and at the 90th percentile of attainment. For measures below 15% weight, the dollar difference between average and excellent performance is trivially small. Then show the operational cost of each measure: data sourcing, calculation time, dispute resolution, and statement complexity. When leadership sees that a 5% weighted measure costs $X to administer and changes zero behavior, the argument makes itself.
Measures should be stable. Change them only when the business strategy genuinely shifts. If you are changing measures annually, reps never build intuition for what matters and spend the first quarter of each year figuring out the new plan instead of selling. Weights can shift moderately year over year (moving from 60/40 to 55/30/15, for example), but the core measures should be consistent for 2-3 year stretches.