How Do I Score My Reps on New Logo Versus Expansion?

The job a two-motion scorecard is actually hired to do

The scorecard exists to solve a specific failure: a single bookings number hides which motion produced it. A rep closes $400K in a quarter. Was that four new accounts at $100K each, or one renewal-plus-seat-expansion on a legacy logo that would have grown whether the rep called or not? The bookings total answers neither question, and a comp plan built on it pays both reps identically while one built the future book and the other harvested the past one.

So the job description of the scorecard is narrow and unglamorous: make the *composition* of a rep's number visible and consequential. Not just reportable — consequential. Plenty of teams already run a report that splits new business from expansion. The report gets looked at during QBR and ignored for the other eleven weeks of the quarter. What changes behavior is when the split feeds a score the rep sees weekly, and when that score is what the manager coaches to and what the compensation plan pays against.

There is a second job, less obvious. The scorecard is an early-warning instrument for pipeline composition. New-logo starvation is a lagging disaster — a team can post three good quarters entirely on expansion while net-new pipeline quietly hollows out, then hit a wall in month ten when the installed base is fully penetrated and there is nothing behind it. Because expansion deals close faster, at higher win rates, with less discovery, a rep optimizing purely for attainment will always drift toward the base. That drift is rational at the individual level and catastrophic at the team level. The two-line matrix catches it roughly two quarters before the revenue line does.

The third job is diagnostic fairness. In most B2B teams, territory quality is wildly uneven — one rep inherits eleven accounts with 400 unpenetrated seats, another gets a clean sheet and a phone. Scoring both against one bookings number rewards the account draw, not the work. Splitting the lines and weighting them lets you set different expectations by segment while keeping one comparable composite. The rep with the fat base has a high expansion bar and a modest new-logo bar; the greenfield rep has the inverse. Both can hit a composite of 4.2, and the number means the same thing.

What the scorecard is *not* hired to do: replace the quota. Quota is a dollar commitment. The composite score is a behavioral picture. They coexist — a rep can hit quota and score a 2.8 because they hit it in one motion, and that gap is the coaching conversation.

Building the matrix: KPIs, weights, and the 1-to-5 levels

Start with the KPI list, not the weights. Six lines is the practical ceiling for a scorecard reps will actually read; four is often enough. A defensible default set for a land-and-expand team:

• New-logo bookings — ACV from accounts with no prior contract. Define "new logo" once, in writing, including how you treat a churned account that returns (most teams call a return after 12+ months a new logo; under 12 months it's a win-back and scores as expansion).
• Expansion bookings — upsell, cross-sell, and seat growth on existing accounts. Decide explicitly whether contractual auto-escalators count. Most teams exclude them, because paying a rep for a price increase written into a contract two years ago is paying for someone else's work.
• Gross retention — this is the guardrail. Without it, an expansion-weighted rep can churn a base account and backfill with upsell elsewhere and look fine.
• Pipeline created — specifically new-logo pipeline created, because this is the leading indicator that fires before the new-logo bookings line goes soft.
• Win rate — keeps volume-based gaming honest.
• Activity or coverage — the weakest line, and the one most prone to gaming. Weight it lightly (5–10%) or leave it off entirely for senior reps.

Then set the weights. Weights are a leadership decision, not a RevOps one — RevOps builds the instrument, leadership aims it. A land-heavy year might run new logo at 40%, expansion at 20%, gross retention at 15%, new-logo pipeline at 15%, win rate at 10%. A base-protection year inverts the top two: expansion 35%, gross retention 25%, new logo 20%, pipeline 10%, win rate 10%. There is no universally correct split. The only hard rule is that the weights sum to 100 and that nobody discovers them for the first time at their review.

The 1-to-5 levels are where most implementations go wrong. A level is not a percentile rank against peers — ranking creates a zero-sum team where helping a colleague costs you. A level is an absolute standard defined against a target. The cleanest mapping:

• Level 1 — under 60% of the line's target
• Level 2 — 60–84%
• Level 3 — 85–100% (this is "doing the job")
• Level 4 — 101–130%
• Level 5 — above 130%

Publish those bands. A rep should be able to compute their own composite on a napkin, and the moment they can't, they stop trusting it. The composite is then simply the sum of (weight × level) across all lines — a rep at level 3 on everything lands at 3.0, which should be the definition of fully meeting expectations, not a disappointment.

Two calibration warnings. First, if your whole team scores 4+, your targets are set too low and the scorecard has stopped discriminating; recalibrate the bands annually against actuals. Second, resist adding a seventh, eighth, ninth KPI. Every line you add dilutes the weight on the lines that matter, and at ten KPIs a rep can be level 1 on new logo and still post a respectable composite — which is precisely the failure the scorecard was built to prevent.

How the scorecard fits into the RevOps stack

The matrix is a layer, not a system of record. It reads from things that already exist and writes into things that already exist, and the implementation work is mostly plumbing on both ends.

Upstream, everything depends on opportunity classification being clean in the CRM. This is the unglamorous prerequisite that sinks most rollouts: if reps hand-select a "Type" picklist value on the opportunity, that field is wrong maybe 15–25% of the time within a quarter — sometimes carelessly, sometimes strategically once the rep learns which value pays better. Derive the classification instead. If the account has a prior closed-won contract, the opportunity is expansion; if it doesn't, it's new logo. Compute that with a formula field or a nightly job against the account's contract history rather than trusting a picklist. Where you genuinely need human judgment — a new division of an existing parent, an acquired subsidiary, a former customer returning after four years — route it through a short exception queue that a RevOps analyst clears weekly, and log the decision so the precedent holds next time.

Downstream, the composite feeds three consumers. Compensation is the loudest: separate accelerators or rates by motion is the standard mechanism, and incentive-compensation platforms exist specifically to administer multi-component plans at scale. Coaching is the most useful: a manager's 1:1 agenda should be driven by the rep's weakest weighted line, which turns a vague "let's talk about your quarter" into "your new-logo pipeline line is a 2, here are the three accounts we're going to work this week." And forecasting is the quietest but arguably highest-leverage: once bookings are cleanly split by motion, your forecast model can apply different win rates and cycle lengths to each, which typically tightens forecast accuracy more than any amount of deal-inspection ceremony, because expansion deals genuinely behave differently — often 2-3x faster to close at materially higher win rates.

A note on where this lives. Teams standardized on a major CRM can host the whole thing in custom objects, formula fields, and a dashboard — the inputs are all already there, you're building the rollup. Teams that want it faster than a CRM project allows start in a spreadsheet, and a well-built spreadsheet is genuinely fine for a team of fifteen: separate KPI columns, a weights row, a SUMPRODUCT for the composite. The spreadsheet's real cost isn't the build, it's the staleness — it works until the quarter someone forgets to refresh it, after which reps stop looking and the instrument is dead. PULSE's free [Pulse Check Matrix](/tools/pulse-check) is this exact model pre-built: define the KPIs, set the weights, score each rep 1-to-5, get one composite Pulse number per rep, no login and no spreadsheet upkeep.

Wiring the score to pay, and what that costs

A scorecard nobody gets paid on is a report. The teeth come from compensation, and there are three common designs.

Differential rates by motion is the simplest: pay a higher commission rate on new-logo ACV than on expansion ACV — a 1.5x to 2x premium on new logo is a common spread for a land-heavy strategy. Easy to explain, easy to administer, and it works. The weakness is that it's blunt: it says "hunt more" without saying how much, and a rep can still ignore new logo entirely if expansion volume is rich enough to clear quota on its own.

Separate quotas per motion fixes that. Give each rep a new-logo number and an expansion number, and pay each independently with its own accelerator. This is the strongest forcing function — you literally cannot clear both without doing both. The cost is administrative: two quotas per rep means twice the quota-setting work, twice the disputes, and real friction when a rep blows out one and misses the other. Many teams add a modest crossover provision (excess attainment in one motion counts partially toward the other, often at 25–50% credit) to keep it from feeling punitive.

Composite-gated bonus is the lightest touch: keep one commission plan, then attach a quarterly or annual bonus gated on the composite score. A rep below a 3.0 composite gets no bonus regardless of dollars closed. This preserves plan simplicity while making the scorecard matter, and it's the easiest to pilot for a quarter without renegotiating everyone's plan mid-year.

On tooling cost, be realistic about the ranges. A CRM you already own is sunk cost — building the scorecard in it is analyst time, typically a few weeks of part-time work plus ongoing maintenance. Dedicated incentive-compensation platforms are almost always custom-quoted and priced per payee; they're worth it when plan complexity or headcount makes spreadsheet comp genuinely risky, and overkill for a team of eight. Lighter commission-tracking tools sit in the low tens of dollars per user per month and often have a free tier — that's usually the right first paid step for a mid-size team that wants attainment visible to reps without an enterprise implementation. Revenue-intelligence and forecasting platforms that automatically segment new business from expansion are custom-quoted and land in enterprise budget territory; they buy automation, not the method. Gamification and scorecard-display tools run cheaper and buy visibility, not rigor.

The honest sequencing advice: build the matrix free first, run it for a quarter without touching comp, see whether the composite actually correlates with the reps you already believe are your best. If it doesn't, your weights or your levels are wrong and you'd have been paying against a broken instrument. Only after it survives that quarter should you wire it to the paycheck. Comp-plan changes are expensive to unwind and reps remember them for years.

Rolling it out without a revolt

The rollout kills more scorecards than the design does. Reps hear "new scoring system" and translate it, usually correctly, as "someone is about to change how I get paid." Three things de-risk it.

Publish the matrix before it counts. Show the KPIs, weights, and level bands at least a full quarter before anything depends on them, and let reps see their own scores in that shadow period. The complaint you want to hear early is "my expansion number is wrong because that deal was misclassified" — that's a data bug, and you want it surfaced while the score is decorative.

Fix classification disputes fast. A single unresolved "that should have been new logo" grievance will poison the instrument for the whole team. Publish the definition, publish the exception process, and turn disputes around in days.

Score the manager, too. A team-level composite for each frontline manager stops the classic dodge where a manager quietly tells the team to chase whatever is easiest this quarter. If the manager's own number reflects both motions, the coaching aligns automatically.

One adjacent trap worth naming: don't let the scorecard silently absorb the customer-success team's job. If CS owns expansion in your model and AEs own new logo, then scoring AEs on expansion creates two owners for one motion and predictable turf conflict at exactly the wrong moment in the account lifecycle. Either give the AE partial-credit weight on expansion sourced through CS, or don't put the line on the AE's card at all and instead weight new-logo pipeline higher. Pick one and be explicit; the failure mode is ambiguity, not either choice.

The same matrix logic travels well outside SaaS, incidentally, which is a decent sanity check on whether you've built it right. A distributor scores reps on new accounts opened versus wallet-share growth in existing ones. A staffing firm splits new client logos from expanded req volume at current clients. An insurance agency separates new policies written from cross-sell into the existing book. The KPI names change; the structure — two motions, explicit weights, absolute levels, one composite — doesn't.

A decision framework for choosing your split

Most of the argument about weights is really an argument about company stage, and it resolves quickly once you name the stage out loud.

Two things that framework encodes and that are worth stating plainly. First, the pilot quarter is not optional — it's the cheapest possible way to discover that your level bands are miscalibrated. Second, the loop back from "composite doesn't match known performers" to "recalibrate" should be expected, not treated as failure. Almost every first pass gets at least one weight wrong.

On cadence: review the weights quarterly as a rhythm, but change them only when the strategy actually changes. A scorecard whose weights move every quarter teaches reps that the target is arbitrary and the correct strategy is to wait it out. The right posture is stable weights with the explicit right to re-weight overnight when the board changes the growth mix — and when you do exercise that right, say so loudly, explain the why, and let reps re-aim deliberately rather than discovering it in a dashboard.

Related questions

Should expansion revenue count toward quota at all?

Yes, but ideally in its own bucket. Blending expansion into a single quota lets reps clear the number without hunting. Separate quotas or a differential rate keeps expansion valued while preserving pressure on new logo.

How do I handle a rep who inherits a huge installed base?

Set segment-specific targets. Their expansion bar should be materially higher than a greenfield rep's, and their new-logo bar lower. The level bands do the normalizing, so both reps' composites stay comparable.

What if customer success owns expansion instead of sales?

Then don't put full expansion weight on the AE card. Give AEs partial credit on CS-sourced expansion or shift that weight to new-logo pipeline, and put expansion on the CS scorecard with a shared gross-retention line.

How long before the scorecard changes behavior?

Expect one quarter of shadow scoring plus one quarter live before behavior shifts measurably. Leading indicators — new-logo pipeline created, new-account meetings booked — move first, usually within four to six weeks of the score becoming consequential.

Can I run this without a CRM overhaul?

Yes. The minimum viable version is a spreadsheet fed by two exports: closed-won by account with a prior-contract flag, and a pipeline snapshot. The CRM work only becomes necessary when manual classification stops scaling.

FAQ

What's the single biggest mistake teams make when scoring new logo versus expansion?

Lumping both into one bookings number. It lets a farmer coast on renewals and a hunter neglect the base, and neither shows up in the data until a quarter goes soft. The fix is structural, not motivational: two weighted lines on one scorecard, with a composite that a rep cannot clear by maxing one and ignoring the other.

How do I decide the actual weights?

The weights are a statement of this year's strategy, so they come from leadership, not RevOps. Name the primary constraint first — too few customers, an underpenetrated base, or churn — and the split follows. Land-heavy years commonly run new logo near 40% with expansion around 20%; base-protection years invert that. Publish them, and make sure they sum to 100.

Can a rep who is excellent at expansion still score well?

Only if they also perform on new logo. Because the composite sums weight × level across every line, a level 5 on expansion paired with a level 1 on new logo produces a mediocre composite — the imbalance is arithmetic, not opinion. That's the point: the score makes the gap visible and gives the rep a specific, unambiguous next move.

What KPIs belong on the card besides the two bookings lines?

Gross retention as the guardrail, new-logo pipeline created as the leading indicator, and win rate to keep volume honest. Activity is optional and should be weighted lightly if included. Stop at six lines — every additional KPI dilutes the weight on the ones that actually drive the strategy.

How often should the weights change?

Review quarterly, change only when strategy changes. Constantly moving weights teach reps the target is arbitrary and reward waiting it out. Reserve the right to re-weight overnight when the board shifts the growth mix, but when you use it, announce it clearly and explain the why so the team re-aims on purpose.

Do I need to buy software to run this?

No. A spreadsheet with KPI columns, a weights row, and a SUMPRODUCT works fine for a small team — its real cost is staleness once someone stops refreshing it. Paid tooling buys automation, comp administration, or visibility, not the method. Build and validate the matrix free first, then add a paid layer only where manual effort is genuinely failing.

Sources

• https://www.salesforce.com/resources/articles/sales-metrics/
• https://hbr.org/2015/04/motivating-salespeople-what-really-works
• https://www.bain.com/insights/topics/customer-loyalty/
• https://openviewpartners.com/blog/net-revenue-retention/
• https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights
• https://www.gartner.com/en/sales/topics/sales-performance-management
• https://www.forrester.com/blogs/category/revenue-operations/
• https://www.saastr.com/category/metrics/
• https://a16z.com/16-startup-metrics/

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)