Price elasticity is not a property of your market

Price elasticity is not a permanent property of your market. It is a response measured under a set of conditions: which customers were observed, what they paid before, how satisfied they were, what the seller changed, and how long the observation lasted. The coefficient is useful only when those conditions travel with it.

That is why a price increase can be accepted by one customer, rejected by another and absorbed by a third without any of the three contradicting the same demand curve. The first question is not “What is our elasticity?” It is “Which mechanism made this buyer sensitive now?”

Why does price elasticity shift when buyer reference points change?

In business-to-business transactions, the price paid last time is often the most immediate reference for the next price. Bruno, Che and Dutta found reference-price effects in both quantity and the transaction price in a large transaction record, and they also found that salespeople carry reference prices of their own. A buyer is not comparing your new quote with an abstract market average. The buyer is comparing it with a remembered event.

The same increase can therefore be a small move from a recent high price, or a large loss against a recent concession. The percentage change is identical. The psychological and commercial state is not. That distinction is already visible in the price increase that is judged before it is paid, where the reception of a change is treated as a question of justification rather than as a fixed property of the product.

Why is customer satisfaction an imperfect buffer against price resistance?

Homburg, Hoyer and Koschate measured reactions to price increases in three experiments. Their result is valuable precisely because it is narrower than the sentence usually built from it: satisfaction weakens the negative effect of the increase on repurchase intentions. Under the tested conditions, the study finds a different response pattern for lower- and higher-satisfaction customers.

The study does not show that satisfied customers will keep buying. Its dependent variable is repurchase intention, the setting is consumer services, and the authors show that the buffer thins as the increase becomes larger. The correct commercial use is therefore diagnostic. Before asking whether a price increase is affordable, identify the part of the base whose satisfaction is already fragile.

That is different from saying “raise prices on loyal customers.” Loyalty can be attached to the firm, to the salesperson or to the value the customer receives. The retention number is measured from your side of the table for a reason: an aggregate renewal rate does not tell you which relationship is carrying the willingness to pay.

How does the quoting process act as an experimental treatment in B2B transactions?

Zhang, Netzer and Ansari model B2B buyers as moving between two latent states: a more relaxed state and a more vigilant state. Their result is not a universal taxonomy of buyers. It is a mechanism inside one longitudinal setting: the seller’s pricing decisions can move buyers between states.

That makes the quote itself part of the treatment. A discount that looks like a one-time acquisition cost can also teach a reference price. A sharp increase that is poorly explained can do more than reduce this month’s quantity. It can move the buyer into a state in which every future quote is read defensively.

The same logic applies on the seller’s side. Bergers and colleagues found that a salesperson’s reference discount shapes the discount granted to a new customer, and that incentives reduce the effect. As a synthesis of the customer-side and seller-side findings, a commercial team can create two reference prices at once: the customer’s memory of what was paid and the seller’s memory of what was given away.

How should commercial teams apply elasticity estimates in strategic decisions?

Use elasticity as a comparison inside a controlled decision, not as a market constant. Compare accounts that share the relevant conditions, keep the reference price visible, and record whether the change was accompanied by a reason the customer could evaluate.

When the outcome is bad, do not immediately conclude that the coefficient was wrong. The coefficient may have been answering a different question. A base with a high share of rep-owned relationships may be sensitive to reassignment, not just to price. A base acquired through heavy discounting may be sensitive to the reference price the seller created, not to the nominal list price. A renewal cohort with a recent service failure may be sensitive to fairness.

The practical test is the same one used in pricing as a positioning decision: what is the price meant to signal, and what does the buyer think it signals? If the answer is “nothing, it is only a number,” the organisation is asking elasticity to carry a positioning problem.

And remember the seller’s side of the record. The discount that outlives the deal shows why an unrecorded concession can become the benchmark for the next negotiation. The price increase is not the first event. It is often the first event the spreadsheet notices.

The final discipline is modest: attach the setting to every elasticity claim. Name the sample, the reference point, the outcome, the time window and the mechanism that changed. If those details cannot be named, the number may still be useful for exploration. It is not ready to steer a price increase.