When repricing is the right problem
Use this framework when
- Discounts, exceptions, or support burden are rising.
- Customers receive materially different value under one flat price.
- The product and customer mix have changed since the last price design.
Do not use it when
- Churn is caused by reliability or weak product value.
- Sales execution, not the price architecture, is the binding constraint.
- You lack the data or customer access to bound migration risk.
A six-part repricing framework
1. Start with economic value, not the current price
Estimate value from the buyer's own arithmetic: revenue created, cost removed, risk reduced, or time returned. The useful question is not "What would you pay?" It is "What changes in your operation when this works, and what is that change worth?" Economic value creates an upper reference point, not an automatic price.
2. Measure willingness to pay through trade-offs
Interviews reveal language and context, but stated enthusiasm is not a demand curve. Use methods that force trade-offs when the decision warrants it. Price-sensitivity questions can establish an acceptable band. Gabor-Granger can test purchase intent at specific price points. Choice-based conjoint can estimate the relative value of features, service levels, and packages. Use the lightest method that can answer the actual pricing question.
3. Choose a price metric customers can understand
A good metric rises with customer value, is hard to game, and is easy to forecast. Per-seat pricing is simple but can punish collaboration. Usage pricing aligns with consumption but creates budget anxiety. Outcome pricing sounds aligned but often creates measurement disputes. Test the metric against both value alignment and the operational cost of billing, quoting, and explaining it.
4. Design tiers and fences deliberately
Packaging should help different customers self-select, not create a maze. Put high-value differentiators where they separate genuine willingness to pay. Avoid withholding basic trust features, such as security or reliability, merely to force an upgrade. Model how the existing customer base maps into the new structure before publishing the structure.
5. Build the migration economics
A headline uplift is not a forecast. Model price, volume, mix, discounts, contraction, churn, support cost, and collection timing. Bound elasticities using your own transaction and renewal history where possible. Run conservative, base, and upside cases, then identify which assumption moves the answer most. The migration decision should be robust to a plausible miss on the most sensitive input.
6. Roll out with explicit trust choices
Choose who moves, when, and with what protection. Options include advance notice, renewal-only changes, time-limited grandfathering, credits, phased increases, or a choice between current and new packages. Explain the value and the logic plainly. Give customer-facing teams a small set of honest answers to the hardest objections, not a script that pretends there is no downside.
A B2B software company replaces one flat plan
A company charges $99 per month to customers ranging from two-person teams to departments with 80 active users. Support hours and workflow volume, not account count, drive cost. Interviews and trade-off research show that small teams value simplicity while larger teams value governance, integrations, and audit history.
The company keeps a simple entry plan, introduces a usage band tied to completed workflows, and places advanced controls in a business tier. Existing customers receive 90 days of notice and can choose a 12-month transition price. The base case expects a 14% increase in recurring revenue, but the decision is approved only because the downside case, including higher contraction and slower conversion, remains cash-positive within the agreed payback period.
Evidence requirements
| Evidence | What it resolves | Important caution |
|---|---|---|
| Invoice and discount history | Pocket price, leakage, and exception patterns | List price alone hides the real system |
| Usage, retention, and support data | Value and cost differences by segment | Correlation does not prove price causality |
| Win, loss, and renewal notes | Where price affects actual choice | Sales notes may encode selective memory |
| Trade-off research | Feature, package, and price sensitivity | Sample and survey design can bias results |
| Migration model | Revenue, churn, contraction, and payback ranges | Elasticity should be bounded, not guessed precisely |
Failure modes that burn trust
Cost-plus storytelling: customers are told about your costs rather than their value. Competitor mimicry: another company's packaging is copied without matching buyer behavior. Grandfathering forever: temporary protection becomes an unmanageable second product. Hidden shrinkflation: limits change without an honest explanation. Average elasticity: one blended assumption conceals that valuable segments react differently. Early panic: normal renewal noise is interpreted as proof that the change failed.
Pre-launch checklist
- The value case is expressed in the customer's economics.
- Willingness to pay is tested through observed behavior or explicit trade-offs.
- The price metric aligns with value and remains predictable.
- Every current customer is mapped to a new package and migration path.
- Price-volume-mix, contraction, churn, and support effects are modeled.
- Notice, grandfathering, and exception rules have owners and expiry dates.
- Monitoring thresholds distinguish normal noise from a material miss.
Limitations
Research cannot perfectly predict behavior under a real invoice. Small samples may not represent the customer base, and competitor reactions can change the market after launch. Price changes also interact with product quality, sales incentives, and macro conditions. Treat the forecast as a range, monitor cohorts by migration path, and preserve the ability to adjust without improvising exceptions account by account.
Related services and cases
Price the value and the migration
We can build a defensible pricing architecture and a rollout that respects customer trust.