The seats you bought and the seats people use are rarely the same number. This guide shows how to count active Salesforce licenses, find the dormant ones, and carry the true down into your renewal.
A Salesforce license count audit reconciles paid seats against active use to find shelfware. This guide walks the scope, the counting method, the traps, and the renewal moves that turn the count into savings.
A Salesforce license count audit is the buyer side check you run on your own org before Salesforce runs one on you. It answers one question. How many seats are you paying for that nobody uses?
Most enterprises cannot answer that on demand. The contract says one number. The org tells a different story once you read it properly.
It checks the distance between what you bought and what you use. That distance is where the renewal savings live. Salesforce sells by user license and by edition, set out on the editions and pricing pages, so the audit reconciles both.
Provisioned seats are what your order form pays for. Active seats are what people actually log into. The gap is the shelfware. Salesforce user license documentation defines the seat types you reconcile against.
Feature access often rides on a permission set license layered on the base seat. These are billable and easy to forget. Count them as their own line, not as part of the base seat.
You pull the data from the org, not from memory. A clean count rests on two reports and one reconciliation.
Use the active users report and the login history, not the headcount your admin remembers. A user can be active in status yet never log in. Both facts matter, and only the report shows them.
Lay the active count beside the order form quantity. The Main Services Agreement governs what you committed to, so reconcile to the signed quantity, not the invoice. Any seat you cannot tie to a working user is a true down candidate.
Salesforce license count reconciliation, illustrative enterprise org
| License type | Paid seats | Active in 90 days | True down candidate |
|---|---|---|---|
| Sales Cloud Enterprise | 400 | 331 | 69 seats |
| Service Cloud Enterprise | 250 | 214 | 36 seats |
| Platform license | 300 | 238 | 62 seats |
| CRM Analytics add on | 120 | 71 | 49 seats |
Three traps quietly keep your paid count higher than your real need. Each one is fixable, and each one shows up in the report.
The standard account team line is that an audit is a compliance exercise, so you only need to prove you are not over deployed. We disagree. In most license counts we ran, the risk was never under licensing. It was the reverse, paying for seats nobody touched and renewing them on autopilot. Treating the audit as a defense against a Salesforce claim misses the real prize. The buyer side move is to run the count as a cost recovery exercise first, find the dormant and over edition seats, and carry that evidence into the renewal as true down leverage rather than waiting to be asked.
Users assigned a higher edition than their role needs are pure overspend. A read mostly user on an Unlimited seat is the classic case. Map roles to the lowest edition that does the job.
Leavers and role changers often keep an active seat for months. Salesforce deactivation guidance frees the seat for reuse. A monthly deactivation routine stops the count creeping back up.
Source: Redress Compliance advisory engagement file, 2024 to 2025.
Count before you renew, not after you sign. The seat you cannot prove someone uses is the seat you should not be paying for.
Four moves turn the count into renewal leverage. Each one needs the report evidence behind it.
It is a buyer side reconciliation of paid seats against genuinely active seats. You compare the contracted quantity to login and usage data to find shelfware. The output is a true down list you can carry into the renewal.
Pull the active user report and the trailing ninety day login history from the org. A user can be active in status but never log in, so use login data, not headcount. Reconcile both against the signed order form quantity.
In our counts the median dormant seat share ran around 21 percent of paid Sales Cloud and Service Cloud seats. Platform and analytics add ons often ran higher. The figure varies by org, so always measure rather than assume.
Run it sixty to ninety days before the renewal. That gives time to deactivate leavers and downgrade editions so the count reflects real need. Counting after you sign locks you into seats you no longer use.
Yes, at renewal you can true down to a lower quantity if your contract allows it and you bring evidence. Mid term reductions are usually blocked, which is why the renewal window matters. The dormant seat list is your leverage.
It is a user holding a higher edition than their work needs, such as a read mostly user on an Unlimited seat. It is pure overspend. Map roles to the lowest edition that does the job before negotiating quantity.
Yes. Permission set licenses are billable feature access layered on the base seat. They are easy to forget and easy to over assign. Count them as their own line and reconcile assignment against need.
Salesforce can review usage against your contract, but the larger buyer risk is paying for unused seats and renewing them on autopilot. Running your own count first protects spend and gives you a stronger renewal position.
The seat reconciliation checklist, the ninety day login test, the edition downgrade map, and the true down evidence template for the renewal.
Used across more than five hundred enterprise engagements. Independent. Buyer side. Built for procurement leaders running the next renewal cycle.
The cheapest license is the one you stop paying for. A clean count turns dormant seats into the strongest true down argument you can take to a Salesforce renewal.
