How to Raise Prices on Existing Customers

TL;DR Raising prices on existing customers starts with a test, and the percentage comes last. Most installed-base increases move the list price and are absorbed by the discount waterfall before they reach an invoice, so realized price on the same cohort before and after is the measure. Undercharging usually begins in the value metric: when the meter stops tracking the value the customer receives, a percentage increase recovers a fraction of the gap and defers the rest by a renewal cycle. Existing customers judge the increase against a held reference price, and it holds when it reads as the price catching up to value already received. Three levers can carry an increase, and four tests at the end route you to the right one.

When a software company decides to raise prices on existing customers, the first question is rarely the right one. The percentage comes last. What comes first is whether the increase will survive the distance between your list price and the invoice a customer receives.

In an ungoverned pricebook, an installed-base price increase is announced, applied to list prices, and then absorbed by the discount waterfall before it reaches a single account. The customers who were already paying below list continue to pay below the new list. Net prices barely move. The company declares success because nobody churned, and the revenue recovery is invisible.

This article separates two problems that the search results conflate: a price level that is wrong, and a value metric that is wrong. Both produce undercharging. Only one of them is fixed by raising a number.

For the procurement dynamics that shape how enterprise accounts receive a price change in the first place, see our enterprise SaaS pricing guide.

The Increase Most Companies Announce Never Reaches Realized Price

What a price increase actually moves

A list price is the top of a negotiation envelope. In a governed pricebook the envelope is narrow and every step down is earned; where discounting is ungoverned, the deal desk simply works downward from it. When a company announces an increase, the announcement moves the ceiling. Whether the floor moves depends on how the discount structure is governed.

Realized price is the net revenue collected per unit, after every discount, concession, and rounding decision has been applied. It is the number that tells you whether the increase did anything. Most companies do not track it on a cohort basis.

Why list-price increases are absorbed before they reach an invoice

In accounts with unstructured discounting, the spread between list price and net price is determined deal by deal, renewal by renewal, without a policy floor. An increase applied to the list price gives the account executive a higher ceiling. The floor does not move unless someone enforces it.

The mechanism is simple. If nothing changes in how discounts are approved, reps who were closing at a habitual discount off the old list will close at a slightly deeper discount off the new list. The net price lands about where it was. The increase funds the next round of concessions. See margin-calibrated discounting for the discipline that prevents this.

The realized price test to run on your last increase

Take the cohort of accounts that renewed in the 12 months before your last list price increase. Pull realized price on those same accounts in the 12 months after. If realized price moved less than the announced percentage, the discount waterfall absorbed the gap.

This test is the opening question for any installed-base repricing exercise. If the answer is “we don’t have that data,” that is itself a finding to surface before picking the next number.

A company we worked with ran a never-discount policy and believed its list moves held. Its reps would not cut price, so they sold a less capable edition and then had license management grant entitlements well beyond what it included. The policy line held on every deal record and the realized price did not, because the approval path tracked the discount and nothing else. The reps had found a lever nobody watched: sell the cheaper edition, then have license management add the entitlements of the edition above it. The customer received the fuller product at the lower price and no discount was ever recorded. A list increase under that arrangement raised the number on the page and changed nothing, because the lever stayed in the reps’ hands and the approval path never looked at it.

We have seen the same shape from the buyer’s side, as a customer of a marketing-automation vendor sold by edition. Customers bought the edition that carried the number of dashboards or active lists they needed, and the rep discounted it down toward the price of the edition below. The reps went out of their way to help customers work around their own company’s pricing.

Undercharging Usually Starts in the Value Metric

What your installed base is paying for today

Undercharging in an installed base is almost always a value metric problem before it is a price level problem. The value metric is the unit the price attaches to: seats, API calls, records processed, revenue managed. When a product’s delivered value grows faster than the metric, the meter falls behind. Customers pay the same or less while receiving more.

A percentage increase applied to a misaligned meter recovers a fraction of the gap. The customer pays more for the same unit. The value they receive through the unmeasured dimension stays unpriced.

What the roadmap has been giving away

Whether an increase holds also depends on what the installed base has been receiving between renewals. Early on, a software company pours every new feature into the base product to close the gap between the price it charges and the value it delivers. All hands are on making the product worth more in the customer’s eyes, and the subscription carries all of it.

As the company matures, that mode tends to persist, because pricing never became a discipline. There is no gate between maintenance and break-fix, which the subscription already owes, and new value that should carry a price of its own. Every release lands in the base product at no charge.

Without that line, a product at scale turns upside down. Investment goes into the product to keep customers from churning rather than to earn a price, and the competitors are doing the same, so the spend buys no position. An increase announced against that roadmap asks customers to pay more for what they have been taught arrives free.

The question to put before any number: of what shipped in the last year, which releases were maintenance owed under the subscription, and which carried a price of their own?

When does raising the price fix the gap, and when does it only defer it?

Does the customer’s bill grow when the customer derives more value from the product?

Where the answer is yes, the meter is working. A price level adjustment on a working meter is a straightforward exercise, bounded by fairness constraints described below.

Where the answer is no, a price increase delays the problem by a renewal cycle. The company collects a modest recovery now and returns to the same conversation at the next renewal, on a metric that has drifted further from the value delivered.

For the decision framework on choosing the unit the price attaches to, see the value metric decision.

How to tell a value metric problem from a price level problem

A value metric problem and a price level problem produce different patterns in the installed base. A price level problem appears as uniform underpricing relative to new-logo rates across Customer Groups with similar usage profiles. A value metric problem appears as a gap between what the meter counts and what customers do: bills that are flat or declining while value from the use of the product has grown.

The invoices read beside the product’s own usage record show which one it has, and no survey is needed. Look at who pays overage: when the accounts at their cap are not the busiest accounts, the unit is capping the wrong thing. And ask whether the seller can see the unit per account per month at all, because a meter the seller cannot observe is hard to price and harder to defend. What to do about the metric is a licensing decision, made above the billing system by people reading more than its ledger.

The two problems can coexist. Separating them comes before choosing an instrument, and both come before picking a percentage.

A client we worked with priced by territory, with the customer’s own regional map setting the count and the fee following it. A new chief revenue officer arrived on the customer side, consolidated the map into fewer regions, and asked for the license fee to fall in proportion. Usage had not fallen and was still growing, and the population running the software was unchanged. Only the label counting them had moved, and the customer held that label. No price-level move would have recovered what that meter was giving away.

Fairness Decides Whether the Increase Holds

Why existing customers judge the same price differently than new ones

Peer-reviewed research on price fairness finds that existing customers hold a reference price their new-logo counterparts do not. The fairness judgment attaches to the relationship, not to the product. A price a new customer accepts without comment draws resistance from an installed account at the same level. The installed account evaluates the change as a loss against a held reference point, where the new customer evaluates it as the starting price.

This is a structural feature of how buyers in ongoing relationships process price changes. No notice period or announcement sequence eliminates the reference price. The question is whether the increase reads as justified within the terms of that relationship.

The escalator is the version of this we see most often in our corpus. On long on-prem maintenance contracts, a CPI clause applied year after year, often on technology the customer never replaced, compounds until the net price sits above what the same vendor would quote a new customer. Maintenance ends up well beyond what a maintenance line usually carries. The customer discovers it at renewal, holding both numbers, and the fairness judgment is made before the vendor has said a word.

What buyers accept when a price moves

Peer-reviewed behavioral pricing research finds that buyers extend fairness credit to increases they read as cost-driven and withhold it from increases they read as demand-driven. An increase framed by rising input costs or infrastructure investment lands differently than one whose rationale is that the market will bear it.

Peer-reviewed research on price fairness also finds that buyers fight a price they read as taking something away from them. They fight it much harder than they fight the same money presented as a discount they no longer get. Customers judge an increase against the price they are used to paying, and a rise from that number reads as a loss. That is what decides whether the increase holds at renewal or starts a re-evaluation of the relationship.

Market Fairness Pricing at renewal

Market Fairness Pricing is the frame that governs how the increase lands. An increase holds when a customer would describe it as the price catching up to value they already receive. It fails when the customer describes it as the vendor extracting from a captive relationship.

The fairness constraint is a pricing constraint. An increase that cannot be explained in terms of value delivered against the value metric is a harder hold than one that can. For the link between fairness perception and commercial outcomes, see price perception in B2B.

We watched a rep under pressure for a lower price answer that it would not be fair to his other customers, who had bought a similar configuration at a similar volume. He then offered to show the buyer the last five contracts, so they could see they were being treated the same as everyone else. The downward pressure stopped. The prospect became a customer and expanded over a relationship that ran for years.

Three Places an Increase Can Land

An installed-base repricing exercise has structural levers. Each changes something different. Understanding the distinction is necessary before choosing an instrument.

Move 1: the price point

The price point is the number attached to the existing value metric. It is the fastest lever to pull and the fastest to be absorbed, because it operates entirely within the existing discount structure.

A price point increase on a well-governed discount structure, attached to a working value metric, in accounts that read the increase as fair, is a clean operation. Remove any of those three conditions and the recovery degrades. The pricebook deviation measure tells you how clean the starting position is.

Move 2: the packaging boundary

The packaging boundary is the line that decides which capabilities sit in which offering, whether the offerings are editions, modules on a platform or add-ons. Moving the boundary changes what a customer is buying, which changes the reference price they hold.

The move assumes that a customer who needs a few more dashboards will step up to the next offering. What we see in practice is that customers buy within a budget. They will not move up a whole package for one capability that sits behind the line. They go without it and fill the gap with a free tool, an internal workaround or a cheaper third party. Packaging geared that finely lowers switching costs, because customers learn to cleave off the lower-value parts of their workflow they do not wish to pay for. It is the lawn service that charges by the bush: a few bushes out back are not worth the price, so you tell him to skip those. The boundary holds when it follows how a Customer Group derives value. Drawn for revenue recovery, it reads as a capability holdback, and the fairness credit goes with it.

Adding a higher optional offering above the base, instead of repricing the base, is a version of this move. It works when the capabilities that define the new offering are ones the target Customer Group already wants. For the relationship between packaging decisions and risk exposure, see pricing risk.

Move 3: the value metric

Changing the value metric changes what the company captures payment for. It is the slowest move to execute and the one with the longest half-life. It is also the only move that closes the structural gap when the meter has drifted from the value it was designed to track.

A metric change on an installed base requires renegotiation of the unit of exchange, not just the number attached to it. The architecture of transitioning existing customers to a new value metric is per-vendor, per-Customer-Group work. Choosing between these three levers is engagement work, not a formula.

Most companies reach for the price point first, and the diagnosis often says otherwise. One company we worked with had research confirming real room in a Customer Group and was ready to raise. Its salesforce had been winning new deals mainly through aggressive discounting, and asking that team to defend higher prices without retraining would have failed. The sequence became: rationalize discounting first, rebuild fluency around value on live deals, then revisit the price once the new habits held. The raise landed only where the field could defend it.

What the Research Says About Proactive Renewal Outreach

The standard recommendation in the search results is to surface pause and downgrade options to accounts that look at-risk before an increase lands. The rationale is that offering flexibility reduces churn by giving customers a path other than cancellation.

Large randomized field experiments on subscription customers found the opposite. Proactively recommending plan changes to existing customers raised churn rather than reducing it. The effect persisted among customers who declined the offer. The contact itself lowered the barrier to reconsidering the relationship.

Why offering options can move the wrong direction

The mechanism is not mysterious. An unsolicited invitation to reconsider a subscription is a prompt to evaluate whether the subscription should continue. For customers in the middle of that evaluation, the prompt accelerates the decision. For customers who were not in that evaluation, it opens one.

The harm concentrated in accounts with erratic usage and declining trends. A health score flags those accounts as at-risk. The intervention reached them at the moment they were most likely to act on a reconsideration prompt.

Risk scores do not predict who responds to outreach

Separate peer-reviewed field experiments found that churn risk and responsiveness to a retention intervention are effectively unrelated. Accounts targeted by risk score are not more likely to respond to an intervention than accounts targeted at random. The effort concentrates on the accounts the intervention cannot move.

Neither of these findings argues against customer communication. They argue against treating a proactive plan-change offer as a risk-reduction instrument. Surfacing an exit is a decision, and it should be made deliberately rather than as a default response to a health score threshold.

Tests to Run Before You Raise Prices on Existing Customers

Before selecting an increase level or a lever, four tests bound the decision space.

What did realized price do after your last increase, on the same cohort? If the answer is unknown, the discount structure is ungoverned and any new increase operates on the same terms.

Which Customer Groups in your installed base are paying for a unit that stopped tracking the value they receive? The answer to this question determines whether a price point increase recovers the gap or whether a metric change is the necessary move.

Where is your installed base priced relative to what a new logo signs today, and can you explain the gap in a sentence a customer would accept? A gap that cannot be explained in value terms is a fairness liability at the next renewal.

If the increase lands and every account renews, what have you learned about whether the number was right? Successful churn avoidance is evidence that the accounts stayed, and nothing more. Net price recovery is the measure.

These four tests will surface a diagnosis. The diagnosis determines the lever. If you want a second read on what your installed base is paying and where the gap lives, talk to a pricing expert.

FAQs