Blog
FSE versus headcount pricing
FSE pricing licenses on a defined worker unit while headcount pricing counts every worker record, and the two can produce very different bills for the same organisation. In a Workday style deal, pin down exactly which workers are counted, exclude inactive and non using records, and lock the unit price so the count cannot quietly inflate.
Key takeaways
- The counting unit, not just the unit price, decides what worker based software costs.
- A broad worker definition can sweep in contingent, seasonal, and terminated records that never use the system.
- Agree the counted population in writing and separate worker types so light users are not priced as full workers.
- Lock the unit price at SKU level with a 3 to 5 percent CPI indexed cap and band the growth rather than leaving it open.
What is the difference between FSE and headcount pricing?
FSE pricing licenses on a defined worker unit, often a full service or full time equivalent measure, while headcount pricing counts every worker record in the system. The two units can produce very different bills for the same organisation, because one normalises to genuine usage and the other counts everything. In a Workday style human capital deal, the first move is to pin down exactly which workers are counted and how, since the unit definition drives the number more than the headline rate.
This is a buyer side distinction worth slowing down on. Workday licenses by worker types and modules, and a contract that prices on raw headcount rather than a usage aligned unit will bill you for people who never sign in. The unit is negotiable, so treat the counting basis as a term to settle, not a fact to accept.
Why does the worker count inflate a SaaS bill?
Because a broad definition of worker can sweep in contingent staff, seasonal workers, terminated records, and non employees who never touch the system, and each counted worker carries a unit price. A generous counting basis raises the bill without raising the value delivered, which is exactly why the definition belongs in the contract rather than in the vendor sizing spreadsheet. Left undefined, the count drifts upward with every record added, whether or not that record represents a real user.
The pattern mirrors the wider 2026 shift toward usage and worker meters, where the buyer carries the risk of an expanding base. Pricing is moving from simple seats toward usage, agent, and outcome models, and worker based licensing is the human capital version of the same dynamic. The protection is the same too: define the unit, bound the growth, and lock the rate.
Which workers should actually be counted?
Only the workers who use the relevant capability should carry a full unit price. Different populations use the system differently, so a single flat count over prices the light users and under serves the buyer. Separate the types and price each to its real footprint.
| Worker population | Typical usage | The buyer move |
|---|---|---|
| Active employees | Full use of the modules licensed | Count as full workers and right size the modules to their roles |
| Contingent and seasonal | Partial or temporary use | Negotiate a separate, lower unit or exclude where they do not use the system |
| Terminated records | No active use, retained for records | Exclude from the counted population by contract definition |
| Non using records | Held in the system but never sign in | Carve out so they are not priced as full workers |
How do you right size worker based pricing?
You agree the counted population in writing, exclude inactive and non using records, separate worker types so contingent and seasonal staff are not priced as full workers, and tie growth to a banded model rather than an open per worker charge. Then lock the unit price at SKU level with a 3 to 5 percent CPI indexed cap so neither the rate nor the basis can drift mid term. Each step removes a way the bill can grow without value behind it.
The terms that bound the count
Five terms keep worker based pricing honest across the term.
| Term | Why it matters |
|---|---|
| Defined counted population | Fixes exactly which workers are billable so the basis cannot expand quietly |
| Worker type separation | Prices contingent and seasonal staff to their real usage, not as full workers |
| Banded growth model | Sets price steps for growth rather than an open ended per worker charge |
| Reduction rights | Lets the count fall when headcount falls, not only rise when it grows |
| SKU level lock and CPI cap | Holds the unit price and bounds the annual uplift at 3 to 5 percent |
What does a right sized worker deal look like?
It looks like a count that matches reality, bounded on both sides. Consider an indicative example: a vendor sizes a renewal on a total worker record count well above the active workforce, including terminated and contingent records. The buyer defines the counted population to active users, prices contingent staff on a separate lower unit, and adds reduction rights so a future restructuring lowers the bill rather than stranding paid units. The committed spend drops to match genuine usage, and a banded growth model with a CPI indexed cap keeps the next term predictable. These figures are indicative and shown to illustrate the mechanics.
What is the move on worker based pricing?
Negotiate the unit before the rate. Define the counted population, separate the worker types, band the growth, secure reduction rights, and lock the unit price at SKU level with a CPI indexed cap. The same discipline applies across worker and usage meters throughout the portfolio, and the full buyer side method is in the SaaS Negotiation Guide. When you want it run for you, our Workday negotiation team sits on your side of the table.
Right size the worker count.
Read the SaaS Negotiation Guide for the full playbook, manage the project cost in Workday implementation costs in the deal, and build leverage with Workday versus the alternatives. To run it with specialists, see our Workday negotiation service.
Download guide →Published market figures reflect 2026 SaaS pricing analyses and are labelled indicative where appropriate.