Pricing software: subscription, usage, or performance-based,
Who bears the risk?

Paying for access, usage, or results does not distribute risk in the same way. A subscription leaves the risk with the customer; usage shifts the budgetary risk; and results share the risk, provided the gain can be measured.

When it comes to pricing, it’s difficult to attribute margin gains to the tool, since your teams retain the final say. Regardless of the model, you need to establish the metrics, safeguards, and exit conditions from the outset.

Subscriptions, pay-as-you-go, and performance-based pricing: software pricing providers no longer all charge the same way. And AI is accelerating these changes.

This article is intended for pricing, finance, and procurement departments that are comparing proposals. It shows that the billing model is, first and foremost, a risk-sharing arrangement, and explains what to check before signing.

A two-pan scale on a desk, with one pan loaded with coins and the other with a printed graph

A billing model is also a way to share risk

When comparing pricing software, we're comparing amounts. But behind those amounts lies a mechanism: we pay for access, usage, or a result.

These three payment methods do not entail the same obligations. They determine who bears the risk if the solution falls short, and when the value is recognized and paid for.

They also influence behavior on both sides of the contract. A publisher does not promote the same things depending on whether it is paid to be used, to be consumed, or to generate a profit.

The billing model is therefore not just a minor detail of the purchase. It shapes the relationship you will have with your publisher for several years.

Subscriptions: Predictable Costs, Risk Bearing by the Customer

Subscriptions remain the most common model. You pay a fixed amount—often annually—to access the solution within a defined scope.

Its advantage is predictability. The budget is known, the finance department can plan for it, and the cost does not vary based on usage.

The downside is simple: if the tool isn't widely adopted, you still have to pay. The risk of underperformance falls largely on the customer.

From the publisher's perspective, the incentive is tied to renewal. A responsible publisher therefore invests in adoption, because a tool that is rarely used will not be renewed.

But this effort still needs to be documented. Training, rule configuration, and monitoring during the first few months: change management plays just as big a role as the license itself in determining the true value of a subscription.

The practice: paying for what you use, at the risk of using less

Pay-as-you-go billing ties the price to volume: the number of SKUs, stores, competitor price checks, or recalculations. The more you use, the more you pay.

On paper, the model makes sense. It prevents you from paying for capacity you don't use, and it lets you start small.

In terms of pricing, however, it has an unintended consequence. When every item tracked or every reading costs money, there is a temptation to narrow the scope to stay within budget.

As a result, we track fewer products, and less frequently. Yet it is often the products we leave out that create invisible price differences, as we explain when discussing the products that are truly worth monitoring.

This practice also shifts the financial risk to the customer. A surge in activity, a promotional period, or an expansion of the product line can cause costs to spike without warning.

Before signing, be sure to ask which unit is billed, how it’s calculated, and what happens if you exceed the limit. A cap or a committed volume can significantly change how you interpret the quote.

The result: shared risk—provided you know how to measure it

Performance-based billing means tying the price to a specific gain: a portion of the additional margin, a realized savings, or a achieved goal. This is also referred to as a success fee.

The idea is appealing because it seems to put everyone on the same page. The publisher is paid only if he creates value, and the client pays only for what he earns.

15%

Only [number] of CIOs prefer to pay for their AI applications based on results, compared to 39% who pay based on usage. They cite results that are poorly aligned with objectives, unpredictable costs, and the difficulty of attributing the benefits.
(a16z, survey of 100 CIOs at large companies (15 sectors), June 2025, figures cited by L.E.K. Consulting)

The challenge lies in measurement. For a result to be billable, it must be observable, attributable to the solution, and defined unambiguously in the contract.

In pricing, the margin is observable. But it also fluctuates based on purchasing negotiations, inflation, competitors' decisions, the weather, and the product mix.

Most importantly, pricing software doesn't make decisions on its own. Your teams validate, adjust, or reject its recommendations. Determining what can be attributed to the tool requires a solid benchmark for comparison—such as when measuring the incremental impact of a promotion.

An incentive that could backfire

A publisher paid based on the measured margin has an incentive to increase it quickly. However, a short-term increase in margin can come at the cost of volume, price perception, or customer loyalty.

This risk is not merely theoretical: it depends onthe flexibility of your products and the role each one plays in your offering. A performance-based contract must therefore include safeguards—not just a percentage.

This approach spreads the risk, but it adds a hidden cost: the cost of measurement, monitoring, and disagreements. This cost must be factored into the comparison.

Three models, three risk allocations

No model is inherently good or bad. Each is suited to a specific situation, and many publishers combine a fixed portion with a variable portion.

ModelWhat You PayWho Bears the Risk?What the model encouragesPoint to Watch For
SubscriptionOne access, for a specific scope and durationEspecially the customer, if they don't use the tool very oftenRenewal, and therefore adoptionWhat the Support Covers
UsageOne volume: references, stores, statements, recalculationsThe client, based on their budgetConsumptionThe temptation to narrow the scope of monitoring
ResultA portion of a measured gainMixed, if the measure is soundThe measured gain, sometimes in the short termThe Measurement Standard and Safeguards

The right choice depends on your level of maturity. A startup needs budget transparency. An organization that can already measure its margin growth can discuss a variable component.

In any case, the question of return on investment comes before the model. We addressed this in our article on the ROI of a pricing solution.

How AI Is Changing Invoicing

AI has introduced costs that vary depending on usage. Every recalculation, every call to a language model, and every simulation consumes computing power, and you pay for that computing power every month.

10%

Some software vendors that charge for AI on a pay-as-you-go basis base their pricing on the results achieved. Others charge based on the effort expended (35%) or the volume produced (55%).
(Bain & Company, “AI Pricing: A Reality Check on Effort, Usage, and Outcomes,” August 2026)

Many publishers therefore pass on these costs through credits or counters added to the subscription. With a pricing model that benefits from being recalculated frequently, the issue becomes a practical one.

If you recalculate your prices daily instead of weekly, does your bill change? And where are the models processed—at the publisher’s, in a third-party cloud, or on your own premises? We go into detail on this topic in our article onon-premises vs. cloud AI.

The rise of agents is also reviving performance-based billing. An agent acting alone seems easy to pay on a per-task basis. But the more a system acts on its own, the more it is necessary to define what it is authorized to decide, asagentic pricing demonstrates.

Questions to Ask Before Signing

These questions apply to all publishers, including Booper. They complement the questions in our series on choosing a publisher, starting with the issue of reliance on an all-in-one publisher.

Question to AskWhat the answer should show
Which unit is billed, and how is it calculated?A written definition that you can verify yourself
What happens if we expand the scope or recalculate more often?A schedule known in advance, a budget cap, or a committed amount
Which AI costs are included, and which ones are billed separately?Credits or meters as described, with their unit prices
If a portion depends on the result, what is the basis for measurement?A comparison period, indicators, and a method established before the start
Who makes the final decision in the event of a disagreement over the measured gain?A written procedure, not case-by-case arbitration
What safeguards are in place to govern price recommendations?Benchmarks for price, volume, and compliance
Our Position
Margin gains are measured collectively before being discussed

Pricing software doesn't make decisions on its own: your teams approve prices, your buyers negotiate, and your competitors make moves. Attributing a percentage of the profit to the tool sparks a debate that can end up taking up more space than the prices themselves.

Regardless of the model chosen, we prefer to work with the client to establish the metrics, the reference period, and the measurement method as soon as proof of value is demonstrated. Progress is then tracked over time, based on criteria agreed upon by all parties.

The billing model should also be considered in light of the contract term. A long-term contract based on results and a short-term subscription-based contract carry different risks.

The gap widens if the publisher changes course. We demonstrated this using a publisher's roadmap as an example.

Finally, think about the exit strategy. Regardless of the model, your rules, history, and product matching should be retrievable, as we explain in the section on reversibility.

The question to keep in mind
If the solution doesn't live up to its promises, who will pay the price, and how will we know?

A billing template addresses the first half of the question. The contract must address the second half.

To include these questions in your evaluation, refer to our specifications for pricing software. And to assess the financial stability of the software vendor itself, ask : Can it deliver on its promises all the way through?

Would you like to compare pricing solutions from different providers?

Our experts will work with you to determine how each billing model affects your budget, your risk, and your prices.

Let's discuss your choice of publisher

Frequently Asked Questions About Billing for Pricing Software

There are three types. A subscription provides access to the solution for a specific scope and duration in exchange for a fixed fee. Pay-as-you-go billing bases the price on volume: SKUs, stores, price lists, or recalculations.

Performance-based billing ties a portion of the price to a measured gain, such as a share of the additional margin. Many publishers combine a fixed portion with a variable portion, particularly to pass on AI costs.

There is no standard price. The cost depends on the number of SKUs and retail locations, the channels covered, the modules selected, the competitive data to be collected, and the level of support provided.

To compare offers, you need to consider the total cost: licensing, setup, interfaces, training, potential AI costs, and exit terms. The billing model also matters, as it determines how that cost will change based on your usage.

This is a model in which part of the price depends on a benefit derived from the solution: an additional margin, a cost savings, or a goal achieved. It is also referred to as a success fee or performance-based payment.

This model shares the risk between the vendor and the client. However, it assumes that the outcome is observable, attributable to the solution, and unambiguously defined in the contract, with a baseline established before the project begins.

Because profit margins depend on many factors outside the tool’s control: purchasing negotiations, inflation, competitors’ decisions, weather, and product assortment. And because your teams approve, adjust, or reject the recommendations.

Determining the portion of the profit attributable to the software requires a measurement method accepted by both parties. Without it, the contract leaves room for disagreements and can sometimes lead to attempts to quickly increase margins, at the expense of volume or price perception.

It makes sense when the scope is still limited, because it avoids paying for what isn't being used. It allows you to start with a single category or region and then expand.

It becomes less favorable when usage increases or fluctuates significantly—for example, during sales periods. In such cases, there is a risk of reducing the number of items tracked or the frequency of data collection in order to stay within budget.

A budget cap or a fixed budget helps keep costs under control.

AI adds costs that vary depending on usage: every recalculation, every call to a language model, and every simulation consumes computational resources. Many providers pass these costs on through credits or usage meters added to the subscription.

Before signing, ask what AI costs are included, what happens if you adjust your prices more frequently, and where the models are hosted. The rise of chatbots is also reviving performance-based billing, which raises the same questions about measurement and safeguards.

‍

Articles
similar
Three lines of light framing a negotiation area, with a price tag in the center
November 17, 2026
Minimum price: Give the salesperson a range rather than a specific price

The floor price is the price below which a sale cannot be finalized without approval. Together with a target price and a ceiling price, it forms a range that guides the negotiation without locking it in.

The sales rep retains their autonomy, the company maintains its margin, and sales that fall below the threshold become transparent. AI helps set these thresholds by customer segment, provided that the sales rep understands where they come from.

Read the blog post
A light strand connecting a series of price tags to decision points on a frieze
November 16, 2026
Price Traceability: Knowing Who Set This Price, When, and Why

The history shows what price was listed. Traceability shows who decided it, when, on what basis, and with what approval.

It is used to provide evidence during an audit, to justify a price to a supplier, a franchisee, or management, and, above all, to link each decision to its outcome. With AI, it distinguishes between what the machine suggests and what a human decides.

Read the blog post
Two meshing gears forced to rotate in the same direction; a spark at the point of contact
November 16, 2026
Automating a Vague Pricing Policy: AI Amplifies Your Contradictions

A vague pricing policy works as long as a human makes decisions on a case-by-case basis. When automated, it applies a set of priorities that no one has chosen, across the entire catalog and with every recalculation.

Conflicting objectives, implicit rules, and exceptions that have become the norm must be resolved beforehand. With AI, the stakes are even higher: a poorly defined objective is optimized with formidable precision.

Read the blog post
Ready to
 booster
What about your profit margins?

The pricing intelligence solution for retail leaders. Accuracy, speed, and controlled margins.

Let's discuss your pricing challenges
‍
‍