A Seat-Based Vendor Reprices on Consumption While 118 Renewals Loom
Corvane sells a workflow automation platform at $60 per user per month. It has 410 customers averaging 84 seats, and 96% of revenue is subscription. Over eight quarters gross margin fell from 82% to 71%. Three accounts, two of them signed at 35% discounts, generate 44% of all workflow runs and drive most of the incremental compute and support cost. Seat counts have flattened: median seats grew 3% year over year while workflow runs grew 61%. Discounting is uncontrolled: average discount is 21%, and 8 of 12 reps routinely approve discounts above their 15% authority using a competitive reason code that no one reviews. The CFO wants 15% more revenue per account within two quarters. The CRO warns that 118 accounts renew in the next six months and two of the three heavy accounts are already in renewal conversations. Leadership is weighing a new package: a $1,800 monthly platform fee, 50 included seats, 250,000 included runs, $0.02 per run thereafter, and multi-year discounts capped at 12%.
1. Evaluate the proposed package against the current model. Which elements would you keep, change, or test before committing, and what migration path would you set for the 118 renewals in the next two quarters?
Read the worked answer
The direction is right. Seats grew 3% while runs grew 61%, so seat pricing has decoupled from both value and cost-to-serve. The specific numbers, though, would produce renewal shocks.
Price the median account before choosing anything. At 84 seats x $60 the customer pays about $5,040 per month. If that account runs 600k workflows, the new price is $1,800 + (350k x $0.02) = $8,800, a 75% increase. The three heavy accounts would see multiples of that. A 75% step at renewal is a churn event, not a repricing.
What I would change:
- Add a collar: cap any account's year-one increase at 20-25% and phase the remainder across two renewals.
- Model the actual run distribution and price to the 90th percentile, not the average, by raising the included allowance or lowering the per-run rate for the top decile.
- Cut or cap the platform fee for small accounts; $1,800 makes sub-$15k accounts uneconomic and accelerates SMB churn.
- Fix the meter first: define what a run is, audit billing data for 60 days, and put the definition in the contract.
Migration: do not flip 118 renewals at once. Pilot with 20 accounts spanning heavy, median, and light usage for one quarter against a control group, then sequence renewals with heavy accounts early and named-account negotiation, so you learn before the peak.
2. Design discount governance the CRO can defend to reps without stalling deals.
Read the worked answer
Discount governance fails when it is a policy document rather than a decision system. Design for speed and visibility.
- Tier authority to economics, not just percentages: rep to 10%, manager to 15%, deal desk to 22%, CRO above that. Make depth conditional: extra discount is approved only in exchange for something concrete, such as a 36-month term, prepayment, a volume commitment, a reference, or a case study. A discount that buys nothing is the leak.
- Kill the unowned reason code. Every exception needs a structured field: named competitor, documented budget constraint, or multi-year commitment. Review the top codes quarterly by rep and by outcome. Competitive with no named competitor is rejected automatically.
- Run a deal desk with a published service level, same-day for most deals and four hours under $50k, so reps do not route around the process at quarter-end.
- Publish rep-level effective discount alongside win rate. Reps change behavior when they see the 25% discounters do not actually win more.
- Do not claw back commissions on deals that were compliant when signed. Change rules forward only.
Expect a short-term dip in close rates. Track it and hold the line for two quarters, or the organization learns the policy is decorative.
3. What two-quarter evidence would convince you the change worked or did not?
Read the worked answer
Judge against a comparison, not a before-and-after snapshot of the same accounts.
Leading, weeks 2-8:
- Effective discount depth on new and renewing deals by rep; target median down 5 points.
- Metering accuracy: share of accounts with clean, reconciled run data.
- Cycle time and approval service levels; repricing must not slow deals.
Lagging, quarters 1-2:
- Realized revenue per account on the renewal cohort versus a matched non-migrated or prior-year cohort. Target +10-15%, not +15% on every account.
- Gross margin per account. That is the whole point. If revenue rises 12% and compute rises 20%, the repricing failed.
- Churn and down-sell in migrated accounts versus control. A 20-25% cap should hold incremental churn under 3 points; above 5 points the collar is too tight or the value story is missing.
- Run volume per active user, to catch customers engineering around the meter through batching or self-hosting.
- New-logo win rate and competitive loss reasons.
Commit to the decision rule in advance: continue if revenue per account rises and churn stays within 3 points; narrow to the pilot segment if churn exceeds 5; reopen the choice of value metric if runs fall while seats rise.