Between 20 and 35 percent of Platform Plus seats went to users who never opened a custom app

A Platform licence is priced on the assumption that the user opens a custom app. There is a simple test for whether they do, and it is almost never run before the renewal.

Prepared by Redress Compliance · August 17, 2026 · Salesforce advisory. 25 to 35 Salesforce Platform and custom app reviews, 2024 to 2025.

Four patterns account for most of it, and they are all findable from data the estate already holds.

Across roughly 25 to 35 Salesforce Platform and custom app reviews led between 2024 and 2025, edition mismatch was the single largest source of waste. The clearest expression of it: Platform Plus seats assigned to users who never open a custom app, often 20 to 35 percent of the list. A Platform licence is priced on the assumption that the user works inside custom applications. When they do not, the estate is paying a premium for capability that is never invoked, and there is a straightforward test for which users those are that nobody runs.

The reason it goes unmeasured is procedural rather than technical. A renewal asks for a seat count. It does not ask whether the seats are the right edition, and the count is easy to produce from the existing assignment list while the edition question requires app usage data. So the assignment carries forward, the mismatch persists, and it grows with the estate because every new user is provisioned against the same assumption as the last one.

The second pattern runs the opposite way and costs more when it lands. API and custom object limits bind during growth, forcing an unplanned edition jump worth 15 to 25 percent. That is the same problem viewed from the other end: the edition was chosen against an assumption about scale, and the assumption expires without warning. An unplanned jump is expensive not only because of the tier price but because it happens at a moment of no leverage, mid term, under delivery pressure, with the alternative being a capability ceiling in production. Watching API call and custom object trends against the ceiling converts that into a planned negotiation.

Underneath both sits the add on layer, where storage and sandboxes were carried at list price while usage sat below 40 percent of entitlement. Bought once, sized generously, never revisited. The common remedy for all three is the same artefact: a clean entitlement map before renewal typically recovered 18 to 30 percent. Build it early enough to act on, secure a documented uplift cap at or below 5 percent while you are in the contract, and treat the seat count as the last question rather than the first. The renewal sequence sits in the renewal negotiation playbook, the wider estate in hidden costs, and the library in the Salesforce practice.

• Pull custom app open events per user, which is the direct test for the 20 to 35 percent of Platform Plus seats that are on the wrong edition.
• Trend API calls and custom objects against the ceiling, so an edition jump becomes a planned negotiation rather than a mid term emergency worth 15 to 25 percent.
• Compare storage and sandbox consumption against purchased entitlement, where usage sat below 40 percent while the add ons carried at list.
• Build the map before renewal, not during, since it typically recovers 18 to 30 percent and only if there is time to act on what it shows.
• Secure a documented uplift cap at or below 5 percent, because an uncapped uplift compounds against whatever base you end up with.
• Ask the edition question before the seat count question, which is the reverse of how the renewal process presents them.

Across roughly 25 to 35 Salesforce Platform and custom app reviews, edition mismatch was the single largest source of waste:

API and custom object limits hit during growth, forcing an unplanned edition jump worth 15 to 25 percent, at a moment of no leverage and under delivery pressure.

Add on storage and sandboxes were carried at list price while usage sat below 40 percent of entitlement, bought once, sized generously, and never revisited.

Watch the briefing · 4:19Where Salesforce Leverage Comes FromWhy the estate evidence, not the seat count, is what changes the conversation.

What is the biggest source of Platform waste?

Edition mismatch, and the clearest expression is Platform Plus seats assigned to users who never open a custom app, often 20 to 35 percent of the list.

Why does that go unnoticed?

Because a renewal asks for a seat count, not for evidence of use. The count is easy to produce from the assignment list; the edition question requires app usage data that nobody is prompted to pull.

How do we test for it?

Custom app open events per user. It is a direct test, the data already exists in the estate, and it identifies exactly which seats are priced on an assumption that does not hold.

What does the entitlement map recover?

18 to 30 percent typically, where it was built before the renewal. The timing matters because the recovery depends on having room to act on what the map shows.

What are the limit driven edition jumps?

API and custom object limits binding during growth, forcing an unplanned move worth 15 to 25 percent. The edition was chosen against an assumption about scale and the assumption expires without warning.

Why are unplanned jumps so expensive?

Because they happen mid term, under delivery pressure, with the alternative being a capability ceiling in production. That is a moment of no leverage, and it is priced accordingly.

How do we avoid them?

Trend API calls and custom object counts against the ceiling continuously. It converts a mid term emergency into a planned negotiation you can prepare for and time.

What about storage and sandboxes?

They were carried at list price while usage sat below 40 percent of entitlement. Bought once, sized generously, and never revisited, because nothing in the renewal process prompts the check.

What should the uplift cap be?

Documented, and ideally at or below 5 percent. An uncapped uplift compounds against whatever base you finish with, which makes it worth more than most single line concessions.

What order should the renewal questions come in?

Edition before seat count, which is the reverse of how the process presents them. Agreeing a count first fixes the mismatch into the new term at the new price.