Multi-cloud leverage: $8.9M using GCP as the BATNA.

The customer's three-year EDP was up for renewal. AWS's initial renewal proposal preserved the existing discount tier and added a modest ramp credit — effectively a hold-the-line offer. The customer's actual consumption had grown 80 percent over the prior term, and the customer's view was that the renewal should be priced against the new run rate, not the old commitment.

AWS's account team had not moved off the opening position through six weeks of internal procurement conversations. The customer's procurement leadership had concluded that without external leverage, the renewal would close roughly at the opening proposal.

What the customer needed

• A credible alternative platform position that AWS would treat as a real risk
• A negotiation sequence that used the alternative without burning the AWS relationship
• An EDP structure that priced the new run rate, not the prior commitment
• Renewal-cycle protection so the leverage was not lost in the next renegotiation

Multi-cloud leverage only works when the alternative is genuinely credible. AWS account teams have seen every fake counterposition; they discount them instantly. A credible BATNA requires real conversations with the alternative provider, real architectural work, and a real workload identified as the migration candidate.

Phase 1 — Build the BATNA (weeks 1-5)

The customer's video transcoding pipeline was identified as the migration candidate. The workload is large, latency-insensitive, and well-suited to GCP's transcoding economics. We engaged a GCP partner to run a formal architecture review and produce a sized commercial proposal. The proposal landed at a 31 percent reduction against the AWS run rate for the same workload.

We did not approach the AWS account team during these five weeks. The customer's posture was simply “we are evaluating options for the transcoding pipeline.”

Phase 2 — Disclose and reset (weeks 6-9)

The disclosure to AWS was framed as a forecast revision: “Our three-year compute forecast under the current EDP renewal proposal assumes the transcoding workload stays on AWS. Based on the GCP proposal we received last week, the financial case is now in favor of moving it. We would prefer to keep the workload on AWS but the current EDP renewal does not support that decision.”

This is the critical sentence. It is not a threat; it is a budget reality. AWS's response was to escalate the renewal to a more senior pricing approver, and the negotiation opened up. Within two weeks, AWS came back with a materially improved tier, a deeper PPA overlay on the transcoding-related services (CloudFront, S3, MediaConvert, and EC2 transcoding instance families), and a renewal-floor clause.

Phase 3 — Close (weeks 10-14)

The final agreement closed with a tier discount 42 percent better than the opening proposal, $4.2M in additional CloudFront and S3 transfer credit, and a renewal floor protecting the new tier as the minimum for the next cycle. The GCP proposal was preserved on file as a benchmark for future cycles but was not executed.

The restructured EDP renewal produced $8.9M in three-year savings against the trajectory the customer was on under the original AWS renewal proposal. No workloads were migrated to GCP.

Where the savings came from

• Tier improvement — $4.6M from the deeper EDP tier achieved through the BATNA-driven escalation
• Service-specific PPA overlays — $2.9M from new overlays on CloudFront, S3, and MediaConvert that addressed the transcoding-workload economics specifically
• Transfer credit — $1.0M in dedicated CloudFront and outbound transfer credit applied across the three-year term
• Renewal floor — $0.4M in modeled savings from the renewal-floor clause, which protects the new discount level against the next cycle's opening position

What the customer did with the savings

Roughly half of the freed budget funded a new live-event streaming initiative that had been deferred for budget reasons. The remainder offset compute growth from the customer's content production pipeline without requiring an annual budget revision.

The renewal-floor clause is the long-lived change. The next EDP renewal opens in roughly 30 months with the new discount tier already established as the floor — AWS cannot reset back to a worse position. The same multi-cloud BATNA framework will be available for the next cycle.