How Do I Score My AEs on More Than Just Closed-Won?

Building the matrix end to end

The mechanics are simple enough to run in a spreadsheet on a Tuesday afternoon, but the sequence matters, because most teams try to buy a tool before they have decided what a complete AE actually is. No platform can make that decision for you.

Step one: name every outcome. Sit down with sales leadership and RevOps and write out the outcomes a fully-functioning AE produces. A workable starting set is closed-won revenue, self-sourced pipeline created, win rate, average deal size, sales-cycle discipline (are deals moving or rotting in stage three), forecast accuracy, multithreading depth, and post-sale handoff quality. Eight or nine lines is the practical ceiling — past that, reps stop being able to hold the model in their head, and a scorecard nobody can recite is a scorecard nobody chases.

Step two: assign weights. Weights are where strategy becomes visible. If you are protecting a mature book, retention-adjacent lines like handoff quality and expansion carry more. If you are pushing into a new segment, self-sourced pipeline gets heavy. The weights must sum to something you can explain — many teams use percentages totalling 100 so the arithmetic is obvious. Closed-won will usually still be the single largest line at 25 to 35 percent; the argument you are having is about the other 65 to 75 percent.

Step three: define the levels. A 1-to-5 level per KPI only works if each level has a written definition. "Level 3 on forecast accuracy" should mean something specific — for instance, commit-category calls landing within a defined variance band over a rolling quarter — not a manager's mood. Write the level definitions down before the first scoring cycle. This is the step teams skip, and it is the step that determines whether reps trust the number.

Step four: score, composite, publish. Score every AE on every line, compute composite = Σ(weight × level), and publish the whole matrix. Not just the composite — the line-by-line detail, visible to the team. A rep who is level 5 on closed-won and level 1 on pipeline creation should be able to see exactly why their composite trails a steadier peer, and exactly which line to move next.

Step five: wire consequences. A score with no teeth is a report card. The teeth live in one of two places: recognition and coaching cadence, or money. Ideally both.

The loop back to weighting is the part that makes this durable. When you launch a product or move upmarket, you re-weight the matrix and the team re-aims within a week — no all-hands speech required, because the incentive already moved.

Where the single-metric model leaks revenue

A pure closed-won leaderboard is not neutral. It actively teaches a set of behaviors, and most of them cost money somewhere downstream of the rep who gets the applause.

Pipeline debt. The most common leak. An AE who closes a large inherited or inbound deal in Q1 has every rational incentive to spend Q2 harvesting rather than sourcing. The bill arrives two quarters later as a coverage hole nobody saw coming, and by then the fix costs more — you are hiring, discounting, or pulling forward deals that were not ready. Scoring self-sourced pipeline as its own weighted line makes the debt visible while it is still cheap to service.

Forecast rot. When only the close counts, sandbagging and happy-ears both become rational. Sandbagging protects the rep's next quarter; happy-ears protects their standing this quarter. Either way, the forecast that finance, hiring, and inventory decisions run on gets worse. Making forecast accuracy a scored line — with a written variance definition — is one of the cheapest interventions in RevOps, because it costs nothing to measure and it changes what reps say in a pipeline review.

Single-threaded fragility. A deal closed through one champion is a deal that evaporates when that champion changes jobs. Bookings-only scoring cannot see the difference between a six-stakeholder consensus close and a lucky single-thread. Scoring multithreading — stakeholder count and seniority reached, documented in CRM — surfaces the risk before renewal season does.

Handoff cost dumped on CS. This is the leak that hides on someone else's P&L. An AE who oversells scope to hit a number generates an implementation that runs long, a customer success manager who spends their quarter firefighting, and a renewal conversation that starts underwater. The revenue looked great on the bookings report and negative eighteen months later. Weighting handoff quality — even scored subjectively by the receiving CSM — reconnects the AE to the consequence.

Deal-shape drift. Bookings-only scoring is indifferent to how the number was made. A quarter hit entirely through heavy discounting, short terms, and non-standard clauses looks identical to a clean quarter on the leaderboard, but the second one is worth materially more in renewal probability and gross margin. Adding average deal size and discount discipline as scored lines makes the shape of the revenue count, not just the size.

The adjacent version of this problem shows up everywhere revenue touches multiple functions. SDR teams scored purely on meetings booked produce no-show-heavy calendars; CS teams scored purely on logo retention neglect expansion; partner managers scored purely on sourced pipeline stop nurturing the partners who influence deals they never get credit for. The matrix method generalizes — the only thing that changes is which eight lines define the complete job.

Concrete numbers to anchor the model

Specifics beat philosophy here, so here are the ranges practitioners actually work with. Treat them as starting points to argue with, not truths — your segment and motion move all of them.

Weight distribution. A common starting split for a full-cycle AE: closed-won 30 percent, self-sourced pipeline 20, win rate 10, average deal size 10, forecast accuracy 10, sales-cycle discipline 10, multithreading 5, handoff quality 5. That leaves closed-won as the biggest single line — which is correct, it is still the job — while ensuring an AE cannot reach a top composite on it alone. Run the arithmetic: a rep at level 5 on closed-won and level 1 on everything else scores 1.5 + 1.2 = 2.6 out of 5. A rep at level 4 across the board scores 4.0. That gap is the entire point of the exercise.

Pipeline coverage. Most teams target 3x to 4x coverage against quota for a quarter, tightening toward 3x when win rates are high and stable and loosening toward 5x for new segments or unproven motions. Scoring coverage as a rep-level line rather than a team-level dashboard metric is what makes individual AEs feel it.

Self-sourced share. For a full-cycle AE, a common expectation is that 30 to 50 percent of pipeline is self-sourced, with the rest arriving from marketing and SDR. Set the level definitions off your own trailing four quarters rather than an industry number — if your team currently averages 20 percent, level 3 should probably be 25, not 40, or you have built a scorecard everybody fails.

Forecast accuracy. Commit-category variance within plus or minus 10 percent over a rolling quarter is a demanding but achievable level-4 or level-5 definition for a mature team; plus or minus 20 percent is a reasonable level 3 for a team just starting to measure it. Define the measurement window explicitly — accuracy measured at week one of the quarter is a very different exercise than accuracy measured at week ten.

Multithreading. Three or more engaged stakeholders on deals above a defined ACV threshold is a workable level-4 bar for mid-market, and five or more for enterprise. Engaged should mean something checkable — attended a call, replied to a thread — not merely a contact record someone typed in.

Scoring cadence and cost. Re-score monthly inside existing pipeline reviews, which typically adds 10 to 15 minutes per rep rather than creating a new meeting. Re-visit the weights themselves quarterly, or immediately when strategy shifts. Tooling cost, if you buy any, spans a wide band: spreadsheets are free, lighter commission and attainment tools commonly start in the low tens of dollars per user per month, CRM platforms that can host a custom scorecard start around the mid-twenties per user per month, and enterprise incentive-compensation platforms are custom-quoted. Check current vendor pricing directly — it moves.

Ramp adjustment. Do not score a rep in month two against a fully-ramped bar. A practical approach is to hold new AEs to the pipeline-creation and activity lines from day one, phase in win rate and deal size at month four, and phase in forecast accuracy once they have had two full quarters of commit calls to be measured against.

Pitfalls that quietly break the scorecard

Too many lines. Twelve KPIs is not a more precise matrix, it is a matrix nobody remembers. Launch with the five or six that matter most this quarter, prove the model, then extend. A rep should be able to name every line on the scorecard from memory, and to tell you which one is currently costing them the most composite.

Subjective lines scored by feel. Multithreading and handoff quality are the ones most likely to become popularity contests. Fix this by anchoring each to a checkable artifact — stakeholder records in CRM, a handoff document completed to a defined standard, a CSM's structured rating on a fixed rubric. If a line cannot be evidenced, either find evidence for it or take it off the matrix.

Changing the weights mid-quarter without saying so. Re-weighting overnight is a feature when it is announced and explained; it is a betrayal when a rep discovers their composite dropped because the model moved under them. Announce weight changes at a quarter boundary wherever possible, and when urgency forces a mid-quarter change, explain what strategic shift caused it.

Gaming a single line. Every metric is gameable in isolation. Self-sourced pipeline invites junk opportunities created to pad a number. Win rate invites refusing to work hard deals. Cycle discipline invites premature closed-lost to clean the board. The defense is the balanced weight set itself — junk pipeline drags win rate down, cherry-picking drags pipeline creation down — plus a periodic look at whether any line is being optimized in a way the customer would resent. When you see gaming, re-weight rather than lecture; the behavior is a rational response to the model you published.

Scoring without coaching. A composite that appears monthly and does nothing produces cynicism fast. Every scoring cycle should end with each rep knowing the single line they are moving next and what specifically they will do differently. This is the part that converts a measurement exercise into a performance system, and it is entirely free.

Letting the spreadsheet rot. The most common quiet death. Someone inserts a column, a formula breaks, the sheet stops getting updated in a busy quarter, and by month four nobody trusts the numbers. Whether you stay in a sheet or move to tooling, assign a named owner in RevOps and put the refresh on a calendar.

Wiring comp to an untested matrix. Do not put money behind weights you have not modeled. Score the last two closed quarters retroactively with your proposed weights and look at the resulting ranking. If your best rep — the one you would clone — ranks sixth, your weights are wrong, not your rep. Fix the model before it touches a paycheck, because a comp plan that pays the wrong person is very expensive to walk back.

Treating tools as the strategy. Visibility platforms broadcast the score, gamification platforms make it feel alive on the floor, and incentive-compensation platforms put money behind it — but none of them author the definition of a complete AE. That authorship is the highest-leverage thing a revenue leader does, because it silently sets what every rep optimizes for the moment they wake up.

Choosing where the teeth live

Once the matrix exists, the buying decision gets much simpler, because you already know what you need the tool to do. The question is only where you want enforcement to sit: in visibility, in pay, or in both.

Visibility-first is the cheaper and faster path. Publishing the matrix on a shared dashboard, in Slack, or on a floor screen, and reviewing it in a monthly one-on-one, changes behavior more than most leaders expect — reps are competitive and they respond to a public gap. Pay-first is heavier but stronger. When a rep can watch their self-sourced pipeline number move their real-time earnings, the beyond-bookings lines stop feeling like homework. The trade-off is that comp changes are slow to design, slow to reverse, and require the payout math to be provably correct — reps will only chase a KPI in their plan if they trust the calculation.

Two adjacent inputs are worth pulling in. Conversation-intelligence data gives you evidence for the qualitative lines, so multithreading and discovery quality get scored off what actually happened on calls rather than off a manager's impression. And your CRM already holds most of the quantitative inputs — bookings, stage velocity, win rate, forecast category history — so the composite can often be assembled where reps already live rather than in a second system nobody opens.

A short checklist before you commit: does the matrix name every outcome that defines the full role; do you own the weights and can you change them without a vendor ticket; can every rep see their own line-by-line levels; is every subjective line anchored to a checkable artifact; has the model been backtested against two closed quarters; and is there a named owner responsible for keeping it current. If the answer to any of those is no, fix that before spending money on tooling — the tool automates enforcement, it does not supply judgment.

Related questions

Should closed-won still be the heaviest weighted line?

Usually yes. Closed-won is the job, and de-weighting it below roughly 25 percent tends to make the scorecard feel disconnected from reality to reps. The goal is preventing a top composite on that line alone, not demoting revenue.

How do I score AEs on a team with wildly different territories?

Score levels against territory-adjusted expectations rather than raw absolutes. A rep in a thin territory hitting 90 percent of a realistic pipeline target should out-score a rep in a rich one hitting 60 percent, or the matrix just measures territory assignment.

Does this work for SDRs and CSMs too?

Yes — the method generalizes. Replace the eight AE lines with the outcomes that define the complete job for that role: for SDRs, meetings held rather than booked, qualification accuracy, and pipeline that survives stage two; for CSMs, retention plus expansion plus health-score discipline.

How long before the composite changes behavior?

Expect one full quarter before scores stabilize and two before behavior visibly shifts. The first cycle is mostly reps discovering which lines they were quietly ignoring, which is itself useful even before anything is wired to pay.

What if a rep disputes their level?

Good — that means the definitions are being tested. Resolve it against the written level definition and the underlying evidence. If the definition is genuinely ambiguous, that is a model defect to fix for everyone, not a negotiation to settle rep by rep.

FAQ

What is the main benefit of scoring AEs on multiple KPIs instead of just closed-won?

It surfaces the whole job rather than one quarter's outcome. Weighting several lines — pipeline creation, forecast accuracy, deal quality, handoff — rewards the rep who builds a durable, predictable book over the one who closed a large deal and let everything else rot behind them.

How do I choose which KPIs to include?

Pick the outcomes and behaviors that define a complete AE at your company, typically eight or nine lines. A common set: closed-won, self-sourced pipeline, win rate, average deal size, sales-cycle discipline, forecast accuracy, multithreading, and post-sale handoff quality. If a line will not change a decision, leave it off.

How do I set the weights?

Set them with leadership against current strategy, sum them to 100, and backtest against your last two closed quarters before publishing. Weights are the mechanism that lets you re-aim the team when a new product launches or you move upmarket — change the weights and the behavior follows.

Will this punish my top closers?

Not if the weights are sane. A genuinely strong closer usually scores well on win rate, deal size, and cycle discipline too. The reps who drop are the ones riding inherited pipeline while neglecting sourcing and forecast honesty — and that drop is the intended signal, not a side effect.

How do I make sure AEs understand their score?

Publish the full line-by-line matrix, not just the composite, and write down what each 1-to-5 level means before the first scoring cycle. Transparency is what converts the number from a verdict into a map — every rep should leave a review knowing exactly which single line to move next.

How often should I re-score the matrix?

Score monthly inside existing pipeline reviews so the composite stays current. Re-visit the weights quarterly, or immediately when strategy or a new product shifts what a complete AE should be chasing. Announce weight changes at a quarter boundary whenever you can.

Sources

• Harvard Business Review — sales compensation and incentive design: https://hbr.org/topic/subject/sales-compensation
• McKinsey & Company — sales and marketing insights: https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights
• Gartner — sales practice research and terminology: https://www.gartner.com/en/sales
• Salesforce — reports and dashboards documentation: https://help.salesforce.com/s/articleView?id=sf.reports_dashboards.htm
• HubSpot — sales metrics and KPI guidance: https://blog.hubspot.com/sales
• SHRM — incentive and variable pay practices: https://www.shrm.org/topics-tools/topics/compensation
• WorldatWork — sales compensation resources: https://worldatwork.org/
• MIT Sloan Management Review — performance measurement research: https://sloanreview.mit.edu/topic/data-data-everywhere/

Related on PULSE

• [How Many Sales Reps Do I Need to Hire for My Logistics Company?](/knowledge/tl0058)
• [How Many Salespeople Do I Need to Hire for My Car Dealership?](/knowledge/tl0052)
• [How Many Producers Do I Need to Hire for My Insurance Agency to Grow My Book?](/knowledge/tl0015)
• [How Do I Figure Out How Many People to Schedule Each Day and at What Times for My Single Store?](/knowledge/tl0002)
• [How Many Attendants Should I Schedule Each Day at My Car Wash?](/knowledge/tl0067)