You rarely win a lost enterprise account back on price.

When an enterprise client starts moving volume to a cheaper venue, the fee table looks like the whole problem. It almost never is. You win the account back by finding the reason the volume left, then putting a better operating answer on the table, with price as a close, not as the strategy.

I learned that selling digital-asset products to institutions, businesses, and high-net-worth clients. I have won a relationship back after volume had already started leaving. I have also lost one I should have closed more honestly. Both are more useful than a generic retention playbook.

Volume leaves quietly

Enterprise accounts rarely send a breakup email. They diversify, test a second venue, and let volume leak while the contract is still in place. If you wait for a formal cancellation, you are managing a post-mortem.

At Coinbase I made a habit of watching client activity rather than relying on the last quarterly review to tell me the truth. In one case, that monitoring showed a large institutional client shifting volume to a lower-fee competitor. The surface complaint was price. The real risk was that a competitor had become the default for a flow we were supposed to own.

That pattern is not unusual in institutional digital assets. In the 2026 EY-Parthenon and Coinbase Institutional Digital Assets survey of more than 350 institutional investors, 61% said they already use a multi-custodian model, against 36% on a single custodian. (EY) Split infrastructure is normal. It is also how a “still our client” relationship becomes a residual relationship. If you cannot see which flows are leaving, you will negotiate as if you still have the book.

The public fee gap makes the misdiagnosis easy. The World Bank’s Remittance Prices Worldwide data for Q3 2025 found banks were the most expensive remittance provider type, at 14.99%, while the global average for sending $200 was 6.36% and the digital remittances index was 4.59%. (World Bank, RPW Q3 2025) In any market where the incumbent rail is obviously more expensive, clients will shop. That does not mean a matched fee is what brings them back. It means price is the language they use to tell you they have a better option.

What actually won the account back

Once I could see the leakage, the wrong move would have been to open with a discount and hope the volume returned. A discount without a change in the product experience trains the client to leave again the next time a cheaper venue appears.

The work looked more like a resale than a save. I had to re-diagnose the flow that was leaving, put the firm’s execution quality and routing capabilities at the center of the conversation, and only then talk commercial terms. Where we needed to, we adjusted the commercial arrangement and moved the client into a higher-support institutional relationship. The relationship came back. Volume grew.

I would not sell that as a formula. I would sell it as a sequence:

1. Catch the leakage while there is still a relationship to save.
2. Separate the stated objection from the job the competitor is actually winning.
3. Put a specific operating advantage on the table, in this case execution and routing, not a brand appeal.
4. Use price last, as a way to make staying rational, not as the reason to stay.

The upgrade mattered because it changed how the client was treated, not because a new label was applied to the same experience. If the only thing you change is the fee, you have not won the account back. You have rented it for another quarter.

The loss I would handle differently

Not every save is available. A high-volume trader wanted fee relief that sat well above a hard constraint I could not move. I let the conversation stay open longer than the constraint allowed. He left.

The lesson was uncomfortable and useful. Some deals should be closed early as a no. If the ask cannot be met, hoping for an internal exception that will not arrive wastes the client’s time and yours. It also turns a clean constraint into a story about how the firm “wouldn’t try.” I should have put the real limit on the table in the first serious conversation, offered the options that did exist, and accepted the loss if those options were not enough.

Winning back an account and losing one for the right reason are related skills. Both require telling the truth about what the firm can actually do.

What this has to do with selling, not just saving

The same discipline shows up before a relationship is at risk.

Founder-led protocols I helped launch on Base were not won with a generic partnership pitch. They needed a concrete path to go live. A retailer I took live on crypto and NFT payments did not need a second price conversation as much as it needed one owner of the handoff among payments, engineering, and the business. At Fidelity, I reached 120% of my annual new-assets target by staying close to the client’s actual objective, not by defending a product line. At Coinbase I reached 160% of the new-funded-account goal and 133% of the cross-sell goal. Retention and origination rewarded the same habit: diagnose the job, then sell the fit.

I now sell embedded financial technology to banks and fintechs. Multi-stakeholder deals in that world fail in the same way enterprise accounts leak: the champion stays friendly, the economic buyer goes quiet, and a competitor becomes the default for the flow that matters.

A practical save playbook

When a flag goes up (usage down, a competitive mention, a sudden interest in “benchmarking fees”), this is the work I would do in the first week, not the first quarter:

Day one: measure the leakage. Which product, which flow, which segment of activity moved? If you cannot quantify it, you are negotiating on a feeling. The client already has a number.

Conversation one: relitigate the job, not the invoice. Ask what is better in the other venue besides price. Speed of execution, missed fills, reporting, funding friction, support responsiveness, and who actually answers when something breaks will tell you more than a narrow price comparison.

Do not lead with a discount. If you need room on price, you can use it later. Leading with it confirms that the relationship is only a fee table.

Put a specific fix on the table. A better tier, a routing change, a reporting cadence, a named support path, a narrower first use case that actually works. “We’ll try harder” is not a fix.

Surface hard constraints early. If there is a line you cannot cross, say so before the client builds a process around an exception you cannot deliver.

Close with a test, not a speech. Agree what “back” looks like: a flow that returns, a volume threshold, a review date. A win-back without a definition of return is a polite conversation.

Price will be in almost every competitive save. It should not be the strategy. The accounts I have kept, and the ones I have won back, turned on whether I could see the leakage early and sell a better way to get the job done. The ones I lost taught me not to confuse optimism with a plan.