FinTech Negotiates Redis Enterprise Down 38 Percent

Situation

A mid sized financial services firm ran Redis across more than sixty instances. The estate had grown organically: a transactional caching layer in front of the core ledger, session stores for customer facing apps, rate limiting on public APIs, and a long tail of internal services that had adopted Redis because it was easy and open. For years the firm treated Redis as a settled, low cost dependency. No one tracked which version ran where, and no one had read the license closely, because there had never been a reason to.

The exposure and the trigger

As of March 2024, Redis moved from an open source license to a model combining the Redis Source Available License and the Server Side Public License. The firm already paid for Redis Enterprise on part of its estate. At the next renewal, the vendor arrived with a sharply higher quote, structured as if the entire deployment needed top tier commercial terms. The increase was large enough to draw the attention of the CISO and the head of procurement, who asked a simple question that the firm could not yet answer: what do we actually run, and which of it truly needs a commercial license.

That question is the exposure in miniature. Without a usage baseline, the firm had no way to test the quote. The vendor held the only narrative, and the renewal clock was running. The risk was not just the price. It was negotiating blind on infrastructure the business could not afford to disrupt.

Approach

We began with a complete map of every Redis instance, direct and embedded, and the version and license state of each. The map separated the deployments into clear groups: internal caching that sat comfortably inside the source available terms, customer facing uses that warranted a closer look, and a small set where a commercial relationship genuinely added value through enterprise features and support. This was the usage baseline the firm had been missing, produced through our open source license risk assessment.

In parallel we built a credible alternative. We costed a migration of the lower risk workloads to Valkey, the open source fork of Redis, with a realistic engineering estimate and a sequencing plan that protected the high availability the ledger depended on. The point was not to threaten the vendor. It was to give the firm a real choice, so the conversation became a comparison rather than an ultimatum. The fork analysis drew on the method in our pillar on Redis, Elastic and database relicensing.

With the baseline and the alternative in hand, we supported the renewal from the buyer side through our open source remediation advisory, pricing the commercial footprint to the workloads that actually justified it.

Outcome

The firm negotiated the renewal down 38 percent against the opening quote. The final agreement covered only the workloads where enterprise features and support added real value, and the lower risk caching and rate limiting moved onto a planned Valkey path. The high availability of the ledger was never at risk, because the migration was sequenced and reversible. Just as important, the firm came away with a living map of its Redis estate and a baseline it could reuse at the next renewal rather than rebuild under pressure.

Lessons for buyers

The first lesson is that a usage baseline is leverage. A vendor quote built on the assumption that everything needs commercial terms collapses the moment you can show what actually sits inside the license. The second is that a credible alternative changes the room. A costed, sequenced fork plan turns a take it or leave it demand into a negotiation between two real options. The third is timing: the firms that map their estate before the renewal arrives negotiate from evidence, while the firms that wait negotiate under a clock. None of this is legal advice, and the firm relied on its own counsel for interpretation of the license, but the commercial discipline is what reset the price.