A discount still bills you forever.
Negotiating a better rate on seats nobody uses reduces the price of the waste without removing it. The remedies run in a specific order, and the order matters because each one changes the position you hold for the next.
Most optimisation stops at the discount
The default response to a large Salesforce bill is to negotiate the rate. It works, it is measurable, and it leaves the underlying structure untouched — the same seats, the same tiers, the same count, at a slightly lower price, renewing forever. Rate is the last lever, not the first, and pulling it first surrenders the evidence that would have made the others work.
Cheaper waste is still waste
A discount applied to seats nobody uses reduces the cost of the mistake without correcting it. The saving is real and it is a fraction of what removing the seats would have produced.
Compare removal with the full cost of alternatives
Removing an unnecessary purchased licence can reduce cost when the agreement permits it. Include minimum quantities, revised discounts and future access needs in the comparison.
Rate is the last lever
Pull it first and you have spent your leverage on the smallest of the four remedies, entering the negotiation with nothing else established.
The constraints are contractual, not technical
Knowing which seats are recoverable is analysis. Actually recovering them runs into contract minimums, renewal timing and repricing mechanics — the reasons a technically correct recommendation fails commercially.
- 01
Salesforce reductions take effect at renewal. A finding delivered after the window closes is an observation, not a saving, and the calendar is unforgiving.
- 02
Contract minimums cap reduction regardless of what the usage data supports, so the honest number is always the capped one.
- 03
Volume-based discounting means shrinking the seat count can trigger repricing of the remainder, and net recovery is lower than gross seat arithmetic implies.
- 04
Mid-term reclassification is contractually difficult even when it is technically correct, which makes downgrade findings advisory rather than promised.
- 05
Removing a seat from someone who still needs occasional access creates a business problem that outweighs the saving — which is what the portal remedy exists to solve.
- 06
The renewal calendar is fixed and buyer leverage decays across it, so the analysis has to land early enough to be usable.
What the optimisation plan is built from
Optimisation begins from the analysis findings and adds the contractual facts that decide which of them can actually be executed, and when.
- 01Classified seat ledger: keep, release, reclassify, or portal
- 02Contract minimums per SKU, customer-confirmed
- 03Renewal date and the decision window preceding it
- 04Current effective rate per tier against list
- 05Volume-discount structure and its repricing sensitivity
- 06Edition entitlements, including licence classes already included at no cost
- 07Add-on licences releasable immediately without waiting for renewal
- 08Downgrade candidates with the evidence supporting each
- 09Light users whose access needs the portal could serve instead of a seat
- 10Headcount plans that will change seat requirements inside the term
- 11Prior amendment history and what was conceded at each
- 12Which remedies can execute now and which must wait for the window
Four remedies, in order of certainty
Release what entitles nothing
Validate unused add-on assignments and deactivated-user entitlements. Technical unassignment may happen earlier than a reduction in the contracted charge.
Remove what nobody uses
Seats with no human login across the window, each carrying its machine-login contrast. The largest certain category, recoverable at renewal.
Reclassify what is over-specified
Full-CRM users doing Platform-tier work, and feature licences enabled for users doing none of the corresponding activity. Advisory, labelled, and priced separately from the certain findings.
Reprice what remains
Only after the first three is a rate conversation worth having — because now it is a negotiation about the seats you actually need, backed by evidence the account executive has not seen a buyer bring before.
The output is a sequenced plan tied to your renewal calendar: what to release now, what to remove at the window, what to reclassify, and what position to negotiate the remainder from.
- Which add-on licences to release immediately, before any renewal conversation
- Which seats to remove at the window, with evidence attached to each
- Which workflows merit a portal assessment after confirming licensing requirements
- Which downgrades are worth pursuing given mid-term reclassification difficulty
- What the capped, net-of-repricing recoverable figure actually is
- When to open the renewal conversation to hold the most leverage
What you receive
Sequenced remedy plan
Every recommendation ordered by certainty and by execution date, mapped against your renewal window rather than presented as an undifferentiated list.
Renewal position paper
The evidence assembled as a negotiating position: what you need, what you do not, and what the estate should cost — written for the conversation, not for a dashboard.
The dated number
Recoverable value capped at the contract floor, net of estimated repricing, captioned with the date each dollar becomes available.
Zero seats don't
Stage 00 is a 30-minute qualifying call at no cost. If the timing or the estate does not justify an engagement, we say so on that call.