Pricing Sensitivity by Segment

A company raises prices by 5% and expects revenue to rise by 5%, but some customers may buy less, switch products or leave while others scarcely change behaviour. Pricing sensitivity by segment compares how defined customer groups respond to price changes so a single average does not hide those differences.

MIT's pricing lecture notes describe customer segmentation in pricing and McKinsey discusses pricing opportunities and differences across customer groups, but the exact sensitivity must be measured, not assumed from a generic industry elasticity. Define the segment, since contract type, use case, order size or service need may be more useful than a broad demographic label, and avoid a segment that is too small for reliable evidence.

Choose the price measure, because list price, discounted realised price and total price including shipping can differ, so use what customers actually face. Define the response as well, as units sold, conversion, renewal, order frequency and contribution margin answer different questions.

Hold other factors in view, because promotions, stockouts, competitor actions and seasonality can move demand at the same time as price, and test comparable groups, since an experiment with appropriate controls can be more informative than comparing two different months without adjustment. Review elasticity, where percentage change in quantity divided by percentage change in price is a common shorthand, though it depends on the range and context.

Watch small bases too, because a segment with few sales can show extreme percentage swings by chance, so include sample sizes and uncertainty. Separate price from mix, since if high-priced items sell more while low-priced items sell less, the average selling price can rise without any price increase, and include service levels, because a premium segment may pay more because it receives faster delivery or support, so do not attribute all demand differences to price alone.

Look at substitutes, because customers with easy alternatives may respond more strongly than buyers needing a specialised product, and consider contracts, as existing fixed-price agreements can delay the effect of a new list price, so track when changes actually reach buyers. Review competitive response and discount leakage, since a competitor may match an increase or cut price, a short trial is not a guarantee of lasting demand, and a list-price increase that salespeople fully discount may change no realised price, so test transactions, not announcements.

Check margins, because more units at a lower price can still reduce profit if the contribution per unit falls too far, so analyse revenue and contribution together. Assess cannibalisation, since a customer may shift from one product tier to another within the business, and track customer lifetime, because a price change can affect renewal and retention beyond the first transaction.

Avoid unfair assumptions as well, as sensitive attributes can be regulated or inappropriate for personalised pricing, so review legal, privacy and fairness rules before use, and be clear with customers, since a confusing price or hidden fee may harm trust even if a short-term conversion test looks favourable. Use ranges, because sensitivity estimated from one 3% change may not hold for a 20% increase, so do not extrapolate far beyond observed data.

Consider capacity, because if demand exceeds available supply a price change can improve contribution without necessarily changing total units sold, and monitor after launch by keeping cohorts and watching churn, complaints and realised margins, with stop rules for a small pilot. Link the work to decisions: for an owner, segment sensitivity shows where a price change alters demand and margin differently, and the goal is a price and offer structure that customers understand and the business can sustain, built on reliable comparison and respectful, lawful implementation, not merely a high elasticity statistic.