$3.2M saved on a three-year Salesforce renewal.

A renewal that had grown without anyone signing for it.

The client — a regional commercial bank with 14,000 Salesforce users across retail banking, business banking, and wealth — was eight months out from the end of its three-year master subscription agreement. Headline ACV had risen from $6.4M in 2022 to $9.1M by 2025. Of the $2.7M increase, less than 30% was driven by deliberate purchases. The rest was a combination of 7% annual uplifts, two opportunistic mid-term true-ups, and a Data Cloud SKU added during a quarter-end conversation that no procurement signature had ever cleared.

Salesforce's initial renewal proposal continued the trajectory — a further 9% list-aligned uplift, a new Einstein 1 Studio attach across all Service Cloud seats, and a three-year commitment with the same uplift mechanics carried forward. The CRO was being told that the bank's CRM cost was now structural and rising.

Why the proposal was excessive

In our experience across 340+ engagements, Salesforce renewal proposals are built around three patterns: bundled SKUs the customer cannot actually consume, automatic uplift mechanics that compound silently between renewals, and growth ratchets that allow Salesforce to claim a higher run rate the moment the customer expands. All three were present here.

Data Cloud was the clearest example. The customer had been billed for Data Cloud Credits since the prior renewal, but the platform had never been provisioned against a production data source. Telemetry showed exactly zero credits consumed in 22 of 23 trailing months.

Seven months. Four workstreams.

Phase one — entitlement reconciliation

We pulled the full set of order forms, amendments, and addenda from the prior three years and reconstructed the entitlement schedule line-by-line. Five Service Cloud add-ons, four sandbox SKUs, and the Data Cloud bundle had been added through different mechanisms — none was visible in a single document. We produced a one-page schedule that, for the first time, the CFO and the CIO could both sign off on.

Phase two — consumption telemetry

We worked with the platform owner to extract login frequency, feature usage, and API call data across every entitled seat and SKU. 2,800 Service Cloud licences had not been logged into for ninety days. 600 had not been logged into in the lifetime of the contract. Data Cloud credits sat at near-zero consumption. Einstein Activity Capture was active on 4% of eligible seats.

Phase three — proposal construction

Rather than respond to Salesforce's draft, we wrote our own. Three-year term. Flat ACV at $9.1M. Data Cloud removed entirely. 2,800 Service Cloud seats step-down at the second anniversary. Uplift cap of CPI or 3%, whichever is lower. Most-favoured-customer language on Data Cloud reintroduction. Public-sector-style renewal options.

Phase four — counter-cycle timing

We held the proposal until the back half of Salesforce's Q4. By week three of January, the account team had moved from "we cannot accept this structure" to "we may be able to find approval at the right commit level." The signed renewal landed on January 28 with the customer's structure intact, no uplift mechanics carried forward, and Data Cloud removed.

$3.2M. And no uplift carried forward.

• Three-year renewal signed at flat ACV — $9.1M versus the $9.9M Salesforce had proposed in year one alone.
• Data Cloud removed from the contract entirely. Most-favoured-customer pricing locked for any future reintroduction.
• 2,800 Service Cloud seats scheduled for step-down at month 24, with documented reduction mechanics in the order form.
• Uplift cap of CPI or 3% on any mid-term true-up, replacing the previous open-ended 7% mechanic.
• Documented saving of $3.2M over the three-year term, signed off by the bank's vendor management office and external audit.

Why this worked

Salesforce renewals reward customers who arrive with their own model, their own numbers, and their own timing. The bank's previous two renewals had been driven by Salesforce's clock — usage data was gathered the quarter before, the account team's proposal set the frame, and procurement was negotiating against a moving target. This time the customer set the frame, the timing, and the terms. The number followed.

If you are coming up on a Salesforce renewal — particularly one that has carried compounding uplifts or includes Data Cloud, Einstein 1, or Agentforce attaches — the work to model your own counter-proposal should start at least nine months out. The Salesforce Renewal Playbook walks through the mechanics in detail.