A national integrated-delivery healthcare network identified 2,300 active-but-dormant Salesforce licenses across Sales Cloud and Health Cloud. Documented utilization, built the return case, and reclaimed $1.4M at renewal — a 29% net reduction. The same playbook is now run quarterly.
The client was a regional health system that had grown through three acquisitions over five years. Each acquired entity brought a Salesforce footprint of its own: provider-relations CRM, donor-management, and population-health outreach. The seat count had climbed steadily; license utilization had never been measured. Salesforce was treated, in the client's words, as "an essential cost" — meaning a budget line, not a managed estate.
A new VP of Vendor Management arrived with a brief from the CFO: quantify SaaS waste across the top ten vendors and recover it. Salesforce was the largest line item. The renewal was eleven months out — enough runway to build a defensible case, recover the shelfware, and reset the relationship around utilization governance rather than seat growth.
The engagement was scoped explicitly around shelfware quantification and reclaim. Edition mix, ramp obligations, and headline rate were in-scope but secondary. The primary deliverable was a defensible utilization dataset, a return case Salesforce could not credibly dispute, and a quarterly governance process the client would run independently after the engagement closed.
Identifying dormant licenses is straightforward; documenting them in a way that survives Salesforce's pushback is not. The diagnostic produced three categories of evidence per inactive seat.
Category one — never activated. 760 licenses had been provisioned during the most recent acquisition integration and never assigned to a user. They had been carried on the contract for thirteen months. There was no Salesforce defense for this; once documented, they were returned in counter-cycle one.
Category two — assigned but dormant 90+ days. 1,180 licenses were assigned to named users who had not logged in for 90 days or longer. The diagnostic established the user employment status (still employed vs. departed), the role of the user (still requiring CRM access vs. role change), and the manager attestation per group of 50 seats.
Category three — low-usage edition mismatch. 360 licenses sat at Health Cloud tier but had been used only for the underlying Service Cloud Console feature. These were eligible for downgrade to Service Cloud Enterprise rather than full return, producing a credit rather than a returned seat.
The strength of a shelfware return case is in the evidence per seat. A blanket "we have 19% shelfware" assertion will be challenged. A 2,300-row spreadsheet naming each seat, its activation date, its last-login date, the user's employment status, and the manager attestation will not.
Salesforce user records reconciled against HRIS active-employee data and Active Directory provisioning. Three reconciliation passes to produce a high-confidence active-employee universe.
90, 180, and 365-day login frequency pulled per user. Dormancy buckets defined and exposed for manager review.
Each dormant seat tagged to its current manager. Two-week attestation window to confirm or decline retention. Non-response defaulted to return.
Spreadsheet built with one row per dormant seat: provisioning date, last login, employment status, manager attestation, recommended action.
360 seats with low-usage feature mismatch flagged for downgrade rather than return. Per-seat downgrade case built with feature-usage evidence.
The diagnostic dataset and attestation cycle converted into a quarterly internal process. Client now runs the playbook independently between engagements.
| Lever | 3-Year Contribution | Mechanism |
|---|---|---|
| Never-activated seats returned (760) | $520K | No defense available; returned in counter-cycle one with no resistance. |
| Dormant 90+ day seats returned (1,180) | $580K | Per-seat evidence package with manager attestation. 87% return rate; remainder retained with documented rationale. |
| Health Cloud downgrade to Service Cloud EE (360) | $210K | Feature-usage evidence supporting downgrade; per-seat credit rather than return. |
| Multi-year price-cap clause | $70K | YoY uplift capped at 5% replacing open-uplift language. |
| Removal of unrequested Einstein add-on | $40K | Einstein for Health Cloud line item the client had not requested. |
An initial proposal to apply the reclaim retroactively as a credit on the current term was rejected. Salesforce does not credit prior periods for utilization; the recovery applies to the renewal term and forward only. The lever was reframed and the recovery applied to the new run-rate instead.
We were paying for 2,300 seats nobody had logged into. The shock wasn't the number — it was that we'd never measured it. The engagement gave us the recovery, but the quarterly cadence is the real outcome. We won't be back to a 19% shelfware rate.
The Salesforce account team will not concede shelfware on an assertion. They will concede shelfware on a defensible per-seat evidence package. The work is in the evidence — the negotiation is straightforward once the evidence exists.
Login frequency alone is incomplete — a dormant user may be on leave, on a role transition, or recently departed. Reconciling against HRIS-active-employee status is what makes the recovery case credible.
Asking managers to confirm or decline each dormant seat does two things: it produces a defensible decision per seat, and it transfers ownership of the cost decision from IT to the line manager. This makes the recovery sustainable across cycles.
Health Cloud seats used only for the underlying Service Cloud feature were downgraded rather than returned. Downgrade preserves the user record, the data, and the configuration — and recovers per-seat cost differential without the disruption of full return.
The largest value of the engagement was not the $1.4M reclaim — it was the quarterly governance cadence that prevents shelfware from re-accumulating. Buyers should treat the recovery engagement as the foundation for an internal process, not as a one-time event.
Most enterprises carry 15–30% Salesforce shelfware. We quantify it and convert it into negotiated reduction at the next renewal.