Salesforce shelfware — licensed but unused capacity — is the single largest source of recoverable savings in most large Salesforce deployments. Across the 500-plus engagements our advisory has supported, shelfware identification and recovery has been the most consistent contributor to the $420 million in cumulative savings we have delivered for clients. The challenge is not that shelfware exists; the challenge is that organizations frequently underestimate how much of it exists and which specific licenses constitute shelfware. This guide walks through the disciplined identification of Salesforce shelfware in 2026: the categories that matter, the data sources that surface the picture, and the analysis framework that produces actionable findings.
What counts as shelfware
The definition matters because the boundary determines what gets recovered. We use a definition with three categories:
Category 1: Provisioned but unused. Licenses provisioned to users who have never logged in, or who have not logged in within a defined inactivity window (typically 90 days). These are the easiest shelfware to identify and recover.
Category 2: Provisioned but underutilized. Licenses provisioned to users who log in occasionally but who do not exercise the functionality that justifies the license tier. A Sales Cloud Enterprise license used purely for occasional account viewing represents a tier mismatch — the user could be on a much cheaper license. These are tier-optimization opportunities, often the largest category by value.
Category 3: Not provisioned but contracted. Licenses purchased in the contract but never deployed. The customer paid for the licenses but never assigned them to users. These are pure-shelfware and produce immediate recovery opportunity at the next renewal.
The data sources for shelfware identification
The shelfware analysis requires data from several sources:
| Data source | What it provides | How to access |
|---|---|---|
| Login History | User login dates, frequency, source IP | Setup > Identity > Login History |
| Active Users | Provisioned users by license type | Setup > Users |
| Contract entitlements | Total licenses purchased by SKU | Contract documents and Order Management |
| Adoption Analytics | Feature usage by user | Setup > Adoption Analytics (where available) |
| Event Monitoring | Detailed user activity | Event Monitoring license required |
| Field History Tracking | Record modifications by user | Per-object configuration |
| Custom report types | Tailored usage analysis | Custom reports against user activity |
For organizations without Event Monitoring, the shelfware analysis can still be substantial using the freely-available Login History data combined with user-level entitlement information. Event Monitoring provides richer detail but is not strictly necessary for first-cycle shelfware identification.
The 90-day inactivity baseline
The 90-day inactivity baseline is a reasonable starting threshold for Category 1 shelfware. Users who have not logged in for 90 days are unlikely to be regular users of the system, and their licenses are candidates for reclamation. The 90-day threshold accommodates:
- Vacation and leave patterns (most leave is shorter than 90 days)
- Seasonal usage patterns (quarterly users still log in within a 90-day window)
- Role changes (users moved off the platform but not yet deprovisioned)
For more aggressive shelfware identification, a 60-day or 30-day window can be used, particularly for high-volume populations where the population dynamics support tighter thresholds. The threshold should be calibrated against the deployment patterns rather than applied uniformly.
The tier-mismatch analysis
Category 2 shelfware — tier mismatches — is typically the largest category by dollar value. The analysis framework:
Step 1: Identify the functional usage patterns
For each active user, identify what they actually do in the system:
- Do they create or edit opportunities? (Sales Cloud required)
- Do they work cases? (Service Cloud required)
- Do they use custom objects beyond the 10-object limit? (Sales Cloud / Service Cloud or higher tier required)
- Do they use advanced features (forecasting, territory management, advanced reporting)? (Higher tier required)
- Do they only view accounts and contacts? (Chatter Plus may suffice)
- Do they only participate in Chatter? (Chatter Free may suffice)
Step 2: Map to the appropriate license tier
Match each user’s functional usage to the lowest-cost license tier that supports their actual needs. The mapping produces a target license mix.
Step 3: Calculate the savings opportunity
Compare the target license mix to the current license mix. The cost delta is the tier-optimization savings opportunity. For populations of hundreds or thousands of users, the delta can be substantial.
Step 4: Validate the functional analysis
Before acting on the analysis, validate the functional findings with the business leaders. Users sometimes have intermittent needs (annual planning cycles, quarterly forecasting) that the activity-based analysis may miss. The validation ensures the tier reductions do not break business processes.
The not-provisioned shelfware category
Category 3 shelfware — licenses contracted but never deployed — surfaces from the comparison of total contracted licenses against provisioned licenses. The gap is pure shelfware, and the recovery opportunity is immediate.
This category often appears in:
- Multi-year contracts with growth assumptions that did not materialize
- Bundled deals where the customer accepted license additions to access discount on other line items
- Expansion projects that were paused or canceled
- Departments that were restructured or eliminated
- M&A transactions where the acquired entity’s license assumptions did not hold
The recovery is typically straightforward at renewal — the customer simply does not renew the unused capacity. The challenge is that customers sometimes forget about this category and renew at the original contract levels rather than renewing at the actually-used levels.
The shelfware analysis output
The shelfware analysis should produce structured outputs:
Inactive users list. By license type, the users who have not logged in within the inactivity window. The list supports deprovisioning decisions.
Tier-mismatch list. By license type, the users whose functional usage indicates a lower license tier would suffice. The list supports tier-optimization decisions.
Not-provisioned list. By SKU, the contracted-but-not-deployed licenses. The list supports renewal-time non-renewal decisions.
Recovery economics summary. The estimated savings from each category, with confidence intervals based on the data quality.
Risk and validation notes. The findings that need business validation (intermittent users, role-based exceptions) before acting on the analysis.
The renewal-time leverage
The shelfware analysis produces leverage that is most usable at renewal time. The renewal conversation can use the shelfware findings as the basis for:
- Reducing the renewed quantity at the affected SKUs
- Migrating user populations to lower-cost license tiers
- Restructuring the contract mix to align with actual usage
- Trading shelfware reduction for discount on remaining capacity
The leverage works because the customer has demonstrated — through the usage data — that the licenses are not producing value. The Salesforce account team typically prefers to retain the customer at restructured economics rather than lose the entire renewal, which creates negotiating room.
The ongoing shelfware management discipline
The shelfware analysis is not a one-time activity. The ongoing discipline includes:
Quarterly inactive-user review. Surface Category 1 inactive users quarterly, with structured deprovisioning of users who no longer need access.
Annual tier-mix review. Surface Category 2 tier mismatches annually, with structured tier optimization at the next renewal.
Renewal-cycle preparation. Surface Category 3 not-provisioned licenses before each renewal cycle, with explicit decisions about whether to renew the capacity.
New-hire and role-change discipline. Process new hires and role changes against the license catalog so that the appropriate license tier is assigned to each new user, preventing the buildup of mismatches.
Departed-employee deprovisioning. Promptly deprovision users who leave the organization, preventing the buildup of inactive licenses.
The Adoption Analytics capability
Salesforce’s Adoption Analytics capability (where available) provides structured adoption reporting that supports the shelfware analysis. The capability surfaces:
- Active user counts by license type and over time
- Feature usage patterns at the user and population level
- Login frequency distributions
- Adoption trends over time
For organizations with Adoption Analytics, the shelfware analysis can leverage the structured reporting rather than building the analysis from raw login data. For organizations without Adoption Analytics, the analysis can still be substantial using the underlying data sources.
What to verify in the shelfware analysis
- The inactivity threshold matches the deployment usage patterns (90 days as default, adjusted for context).
- The functional usage analysis covers all material capabilities (CRM features, custom objects, advanced features).
- The tier-mismatch findings are validated with business leaders before action.
- The not-provisioned analysis matches contracted entitlements against actual provisioning.
- The recovery economics are calculated at the appropriate negotiated pricing levels.
- The intermittent-user exceptions are identified and protected from over-aggressive deprovisioning.
- The renewal-time strategy is documented before the renewal conversation begins.
- The ongoing management discipline is established to prevent shelfware buildup in future cycles.
Shelfware identification is the foundation of the Salesforce cost optimization conversation. The 34 percent average reduction our advisory secures against Salesforce’s opening renewal positions includes a substantial shelfware component, and many of the largest individual savings stories begin with rigorous shelfware analysis. The discipline of running the analysis annually — not waiting for the renewal to start the conversation — positions customers to negotiate from a position of strength rather than reacting to whatever the Salesforce account team proposes.
For most organizations, the right approach is to treat shelfware as a structural feature of large Salesforce deployments that requires ongoing management rather than as an occasional cleanup exercise. The analysis discipline, the validation rigor, and the renewal-time leverage application together produce material recurring savings that compound across renewal cycles.
The shelfware patterns by product category
Different Salesforce product categories produce different shelfware patterns. The 2026 patterns we typically observe:
Sales Cloud shelfware typically reflects role-based mismatches — users on Enterprise edition who only need Professional, or users on full Sales Cloud who only need Chatter Plus. The shelfware is principally tier-based rather than user-count-based.
Service Cloud shelfware often reflects seasonal staffing patterns. Service operations scale staff up and down, but the license counts often remain at peak levels rather than tracking the actual staffing. The opportunity is in negotiating flexible capacity arrangements that match the actual staffing patterns.
Marketing Cloud shelfware frequently reflects feature-set mismatches — customers who purchased Marketing Cloud Studio bundles but who only use a subset of the bundled capabilities. The shelfware is in the unused Studios.
Data Cloud shelfware can be substantial when credit consumption is below the committed minimums. The committed credits become shelfware when the actual data ingestion, identity resolution, and segmentation volume falls short.
Einstein AI shelfware typically appears as unused AI features and unrealized prediction adoption. The license is purchased on a forward-looking basis but the deployment does not produce the expected feature exercise.
MuleSoft shelfware appears in unused integration capacity (vCores, API call allocations) where the integration volume came in below expectations.
Tableau shelfware often reflects tier mismatches (Creator licenses for users who need only Viewer capabilities) and inactive users.
Slack shelfware can appear in inactive guest users, deprovisioned-but-licensed accounts, and tier mismatches between Business+ and Enterprise Grid.
The pattern recognition by product helps focus the analysis — different products warrant different analytical approaches.
The shelfware reporting framework
Effective shelfware reporting follows a structured framework:
Executive summary
A one-page summary identifying total shelfware exposure, the breakdown by category, and the prioritized recovery opportunities. The summary supports executive decision-making about renewal strategy and resource allocation.
Detail by license type
For each license SKU, the detail showing provisioned count, active count, recommended count, and the cost delta. The detail supports the specific renewal conversation by SKU.
Detail by user population
For each user population (department, role, location), the detail showing license assignments, activity patterns, and tier appropriateness. The detail supports population-level decisions.
Specific user list
The user-by-user list of inactive users, tier-mismatch users, and other categories. The detail supports the specific deprovisioning and retier decisions.
Risk and validation notes
The exceptions, edge cases, and validation requirements that prevent the analysis from being applied blindly. The notes ensure that operational realities (intermittent users, seasonal patterns, role transitions) are respected.
Renewal-strategy recommendations
The specific recommendations for the renewal conversation — which line items to reduce, which to non-renew, which to retier. The recommendations translate the analysis into action.
The data quality challenges
Shelfware analysis is only as good as the underlying data. Common data quality challenges:
Frozen user accounts. Users marked as frozen or deactivated may not appear in standard activity reports, producing artificial "inactive" findings. The analysis should validate the user-status filtering.
Service accounts and integration users. Some users are not human users but integration accounts. The analysis should exclude these from the inactive-user identification.
Time zone effects. Login data may not consistently reflect user time zones, producing artifacts in time-based analyses.
Multi-org environments. Customers with multiple Salesforce orgs need to consolidate the analysis across orgs to produce accurate findings.
Sandbox license effects. Sandbox licenses sometimes appear in production-org reporting in confusing ways. The analysis should explicitly separate production from sandbox license accounting.
Recently provisioned users. Users provisioned in the last 30 or 60 days may not yet have established activity patterns. The analysis should give appropriate latitude for new users.
The role-based archetypes for license assignment
Beyond the activity-based analysis, role-based archetypes help structure the license assignment discipline. The common archetypes:
Power user. Heavy daily usage, custom report authoring, advanced feature exercise. Justifies Enterprise or Unlimited edition.
Daily operator. Regular daily usage of core CRM features (opportunities, cases, leads). Justifies Enterprise edition.
Periodic contributor. Less frequent usage focused on specific workflows. Often fits Professional edition.
Manager and reviewer. Dashboard and report viewing, occasional record review. Often fits Professional or higher with strong analytics emphasis.
Executive viewer. Dashboard viewing, occasional account or opportunity review. Frequently over-licensed; often fits Chatter Plus or Platform license.
Field worker. Mobile-centric usage with specific transactional patterns. License tier depends on functional needs.
Casual user. Infrequent access for specific tasks. Often fits Chatter Plus or lower tier.
The archetype mapping supports systematic license assignment that matches functional reality rather than defaulting to uniform tiers across the organization.
The cumulative shelfware effect across renewal cycles
Shelfware tends to accumulate over renewal cycles unless actively managed. The accumulation patterns:
Growth-cycle accumulation. Each growth purchase adds licenses; subsequent contraction does not always trigger right-sizing. Over multiple cycles, the licensed quantity drifts above actual usage.
Bundle accumulation. Bundle purchases sometimes include licenses that produce no incremental value but that come bundled with desired purchases. The bundled licenses become shelfware that never gets addressed.
Acquisition accumulation. M&A activity often produces overlapping license footprints. The post-acquisition rationalization frequently leaves shelfware unaddressed.
Reorganization accumulation. Internal reorganizations move users between roles but do not always trigger license tier reassignment. Over time, the tier mismatches accumulate.
Feature-launch accumulation. New Salesforce features sometimes come with new license requirements. Customers add the new licenses but do not always achieve the expected deployment, producing shelfware.
The active shelfware management practice catches these accumulation patterns at each cycle rather than allowing them to compound. The compound savings from disciplined shelfware management over multiple cycles can be substantial — often the largest single source of Salesforce cost optimization.