Reversibility: What do you get if you switch to a different pricing plan?
October 9, 2026
A reversibility plan describes what a customer recovers when leaving their service provider, in what format, within what timeframe, and at what cost. For a pricing solution, four assets are involved: raw data, rules, arbitrage history, and the product matching table. Generally, only the first of these is covered, while the other three must be reconstructed manually, a process that takes months.
Reversibility is negotiated before signing, based on three principles: specifying the assets to be returned, requiring a format that is readable without the tool, and quantifying the timeline, cost, and parallel operation. The goal is not to leave, but to have the option to do so.
A selection of pricing solutions evaluates features, the model, customer references, price, and timeline. It very rarely considers the release date.
It's understandable: when it's time to sign, no one wants to talk about breaking up. The problem is that this is the only time when it makes sense to bring up the issue.
Once you've signed, you no longer negotiate your exit—you find out what it is.

What is a reversibility plan?
A reversibility plan is the part of the contract that describes how a customer recovers what they have entrusted to or accumulated with a service provider, and how they can switch to another solution without any interruption in service.
It specifies the scope of the handover, the format, the timeline, the cost, and the period during which the old and new systems will run in parallel.
When it comes to pricing solutions, they rarely cover the essentials. Your raw data belongs to you, and no reputable vendor will withhold it. But you had that data before you signed the contract. What you’ve built over three years of use is something else entirely.
Why Everything Comes Down to the Moment Before the Signing
Before signing the contract, your bargaining power is at its peak, and walking away costs nothing. Once the system goes live, you have trained teams, well-established processes, and a genuine dependency.
Asking the question up front isn't a sign of mistrust. A publisher who gives a clear answer about the ability to switch platforms demonstrates confidence in its value; a publisher who dodges the question suggests that part of its customer retention strategy relies on how difficult it is to leave.
The four assets you create with your publisher
Raw data: sales, prices charged, inventory, and reports. Thisis generally the only type of data covered—and the least critical—since it comes from your systems. The only real question is the format, as with the interfaces described in “Integrating Your Pricing Data Without a Major IT Project.”
The rule set: hundreds of rules, each with its own priorities, exceptions, and scope. An export in a proprietary format—or a list of parameters without the logic that ties them together—is virtually worthless elsewhere.
The history of arbitration decisions: every recommendation that was accepted, modified, or rejected, along with the reason for each decision. This is the company’s record of decisions—its most valuable, yet often overlooked, asset.
Product matching: the table that indicates that your product code corresponds to a specific product code at a particular competitor. It is often treated as the software vendor’s expertise rather than as customer data, and this is the most common source of contractual friction. See our product matching offering.
How Much Does It Cost to Rebuild a Correspondence Table?
Illustrative example: With 20,000 product listings tracked across five competitors, that amounts to up to 100,000 entries that need to be revalidated.
At 30 seconds per check, that comes to about 830 hours—roughly six months of a full-time employee's work. Meanwhile, competing price indices are unreliable.
A conversion table exported on the day of departure is accurate only for that day. Every change a competitor makes to its product line reduces its value slightly.
Therefore, it is not just the table that must be recoverable. It also includes the criteria used to match the products and the history of manual corrections: without them, the table cannot be maintained.
Booper Team, retail pricing since 2014
In the same series: “Managing an AI-Powered Pricing System”
The Gap Between Perception and Reality in Migration
Migrations are not impossible. But policymakers' estimates of them before they actually occur are consistently overly optimistic, to the point of rendering the schedule unusable.
89% of executives believe they can switch AI providers in less than four weeks. Among those who have tried, 58% say the migration failed or required much more effort than expected.
(Zapier, survey of more than 500 U.S. business executives, conducted in early 2026, published in April 2026)
This survey focuses on AI providers in the United States, but the same principle applies to pricing. In the same study, 46% cited data migration as one of the main risks, on par with reliance on a single vendor.
There are three reasons for this discrepancy. The export exists but is unusable: hundreds of thousands of lines with no description of their structure. Implicit knowledge cannot be exported: the reasons behind the parameters, and the workarounds known to users.
And the transition isn't instantaneous. We need to run two systems in parallel, which the old contract must allow for. It's the same constraint as when migrating from Excel to a pricing platform.
What a Contract Must Guarantee
There's no need to pile on the clauses. Three principles make the difference between a theoretical right and a viable solution.
Specify what is being returned. A general statement such as “the customer’s data is returned to them” will be interpreted in the narrowest sense when the time comes. The four assets must be listed, including product matching.
Require a format that is readable without the tool. That is what distinguishes an export from a usable output. An open format, with a description of its structure.
Estimate the timeline, cost, and parallel processing. A number of days rather than an obligation to use specific resources, a non-billed deliverable treated as a separate service, and access to the old system during the migration.
What's the Point of Demanding It?
The models themselves. The trained algorithms are at the heart of the publisher's intellectual property, and their disclosure will never be permitted.
This isn't a shortcoming: a model can be retrained using your data in just a few weeks. A rule set, a history of arbitrage trades, and a matching table can only be rebuilt manually, which takes months.
Test its reversibility before you actually need it
An untested clause is an assumption. Requesting a full report once a year—with no intention of leaving—ensures that the deadline is met, that the format is readable by your teams, and that the scope aligns with the contract.
This request has a second effect: it fosters a balanced relationship. A customer whom we know might leave is a customer with whom we negotiate differently.
What if the contract has already been signed?
Request a trial refund, even if there’s no clause requiring it: a refusal is, in itself, a form of information. Document internally the reasons behind the rules, which effectively allows for reversibility. And include the missing clauses in the next renewal.
If the decision to make a change has already been made, the issue becomes operational: see how to switch pricing tools without repeating past mistakes.
Negotiating the right to switch providers isn’t the same as preparing to leave, any more than taking out insurance is the same as anticipating a claim. A publisher that retains its customers through customer satisfaction builds a more solid business than one that retains them through their data.
See also in this feature
- The Booper Solution: Booper Pricing Software
- Selecting and Implementing a Pricing Tool: From Identifying the Need to Going Live
- Pricing tools: comparison of the best pricing software 2026
- Implementing a Pricing Tool: The Complete 8-Step Method
- POC Pricing: How Can You Verify That a Solution Truly Adds Value?
FAQ
This is the section of a contract that describes how the client retrieves what it has entrusted to or established with a service provider, and how it transitions to another solution without interruption: scope, format, timeline, cost, and the period during which the two solutions operate in parallel.
It is the ability to switch from one publisher to another while retaining, in a reusable format, four assets: raw data, the rule set, the history of decisions, and the product-matching table with competitors.
Generally, only the first one is covered by the contracts.
Three principles are sufficient: identify the returned assets, including product matching; require a format that is readable without the tool; and quantify the timeframe, cost, and period of parallel operation.
Because the export does not include a description of its structure, because implicit knowledge is not exported, and because the migration requires two systems to run in parallel.
According to a Zapier survey published in April 2026, 58% of executives who attempted to switch AI providers say the process failed or required much more effort than expected.
No, and there's no point in asking for it: they are the publisher's intellectual property. A model can be retrained in a matter of weeks using your data, whereas rules, decision-making criteria, and matching tables take months to rebuild.
By requesting a full report once a year, with no intention of leaving: Is the deadline met? Is the format readable without the tool? Does the scope comply with the contract?
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.
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.
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.
