Annual commitment locked 10 to 20 percent of spend in seats that could have flexed down
Workspace looks like simple per user pricing, which is why the commitment term rarely gets examined. The annual plan is cheaper per seat and sets a floor you cannot go below, so a workforce that moves pays for one that does not.
Prepared by Redress Compliance · August 15, 2026 · Google advisory. Based on 20 to 30 Workspace negotiations, 2024 to 2025.
Executive summary
The commitment term is a pricing decision, not an administrative one. Estates on annual commitment locked 10 to 20 percent of spend in seats they could have flexed down.
Seat drift compounds it. Between 8 and 15 percent of paid seats belonged to leavers or dormant accounts nobody reclaimed.
Renewal is the only clean moment to fix either. An annual commitment sets a floor, so mid term cleanup is administrative work with no financial effect.
Edition fit is the second lever and it compounds with the first, since 30 to 50 percent of seats sat on an enterprise edition while consuming business edition features.
The baseline is the first negotiation. Volume and term discounts apply to whatever seat count you bring, so bring the right one.
Annual against flexible, priced honestly
| Dimension | Annual commitment | Flexible plan |
|---|---|---|
| Unit price | Lower per user | Higher per user |
| Seat floor | Cannot drop below the committed count until renewal | Add and drop monthly |
| Mid term adds | Allowed, and co terminate with the commitment | Allowed, priced at the flexible rate |
| Fits | Stable headcount with predictable growth | Volatile headcount, contractors, dated projects |
The comparison that decides it. The flexible premium per user is a known number. The cost of carrying locked seats you cannot release is not, because it depends on attrition and project ends that nobody models at signature. In the estates we reviewed the second number was consistently the larger of the two, to the tune of 10 to 20 percent of spend, and it was invisible because a locked seat produces no alert, no invoice line, and no decision. It simply keeps billing at the cheaper unit price everyone chose it for.
The moves that cut a Workspace renewal
- Reclaim dormant seats and leavers before the seat count is set, since 8 to 15 percent of paid seats had no user behind them.
- Match the commitment to headcount volatility, not to the lower unit price, and price both plans against your real attrition.
- Right size editions so light users sit on a business tier that covers their real need, per the edition mix guide.
- Segment by feature use rather than job title, because the seniority of a user says nothing about whether they touch data regions or advanced controls.
- Trade term for price protection, using a multi year commitment to lock against future increases rather than to buy a deeper headline discount.
- Run the review ninety days out, so reassignment and reclamation are finished before the quote is built.
The Google Workspace negotiation guide
Edition mapping, commitment terms, the annual versus flexible trade, and the buyer side moves across the Workspace estate.
Get the guide →A cheaper unit price on a count you cannot change
Annual commitment is presented as the sensible default: a lower price per user in exchange for a year of certainty. The exchange is real and the certainty runs one way. Google is certain of the revenue. You are committed to a seat count set before you knew who would leave, which project would end, or which acquisition would arrive. In the estates we reviewed that asymmetry cost 10 to 20 percent of spend, held in seats that could have flexed down but could not.
The reason it persists is that a locked seat is silent. An unused licence generates no alert and no incremental invoice, and the total looks correct because it matches the commitment everyone signed. Compare that with the flexible premium, which is visible on every quote and therefore feels like the expensive option. Buyers systematically choose the cost they can see over the cost they cannot, which is a reasonable instinct and the wrong answer whenever headcount actually moves.
Seat drift makes it worse in the same quiet way. Between 8 and 15 percent of paid seats belonged to leavers or dormant accounts, and they survive partly because reclaiming them mid term changes nothing financially. Under an annual commitment you have already agreed the floor, so an admin who deprovisions a departed user has done good hygiene and saved no money. The incentive to clean up only exists in the ninety days before renewal, which is precisely when nobody has time unless it was scheduled.
That is the practical conclusion. Renewal is the only clean moment to change the count, the editions and the plan, so the audience review has to happen before the quote lands rather than in response to it. Pull usage data, segment by what people actually touch, reclaim the dormant accounts, and decide annual against flexible on your real attrition rather than on the unit price. Then negotiate. Volume and term discounts are genuine, but they apply to whatever baseline you bring, and a deeper discount on seats nobody uses is still an overpayment. The edition detail sits in the pricing guide, and the platform comparison in the Workspace against Microsoft 365 brief.
- Seat and edition right sizing from actual usage, with dollar figures per group
- Every risky clause flagged with the exact quote, the page, and the replacement language
- A negotiation playbook, talking points, and a two page executive brief on day one
The ninety day runway
Pull the usage data
Active user and feature usage from the admin console, segmented into light, standard and power groups by what people actually touch.
Reclaim and reassign
Remove leavers and dormant accounts, move light users to the lowest edition that covers their need, and settle the real seat count.
Choose the plan, then negotiate
Model annual against flexible on your attrition, fix the baseline, and take volume and term into the discount conversation.
What the Workspace file shows
Across roughly 20 to 30 Google Workspace negotiations in 2024 and 2025, the count and the plan mattered more than the discount:
On estates that chose annual for the lower unit price without pricing their own attrition against it.
Leavers and dormant accounts carried into the next commitment because the cleanup was never scheduled.
The patterns: the plan chosen on unit price, the seat count inherited from last year, and the audience review attempted after the quote rather than before it.
The buyer side move is to fix the baseline in the ninety days you actually control. The wider library sits in the Google practice.
Your first five moves
- Pull active user and feature usage data from the admin console ninety days before renewal.
- Reclaim every leaver and dormant account before the seat count is fixed, and document the reclaimed total.
- Segment users by what they actually use and map each group to the lowest edition that covers its real need.
- Model annual against flexible on your own headcount volatility, pricing the locked seat cost rather than only the flexible premium.
- Take the right sized baseline and a multi year term offer into the renewal. The Google negotiation service runs the review with you.
Frequently asked questions
What is the difference between annual and flexible commitment?
Annual commitment is cheaper per user but locks your seat count for the year, so you cannot drop below the committed number until renewal. Flexible costs more per user and lets you add and drop monthly. The right choice follows workforce stability rather than unit price.
When does flexible actually cost less?
Whenever the premium per user is smaller than the cost of carrying locked seats you cannot release. Volatile headcount, contractor heavy teams and projects with a known end date frequently fall into that category, and estates on annual commitment locked 10 to 20 percent of spend in seats they could have flexed down.
How many paid seats are usually dormant?
Between 8 and 15 percent in the estates we reviewed, belonging to leavers or accounts nobody reclaimed. They persist because nothing in the billing cycle forces a review, and because reclaiming a seat mid term does not reduce an annual commitment anyway.
Why is renewal the only clean moment to fix the count?
Because an annual commitment sets a seat floor you cannot go below until it ends. Mid term additions are allowed and co terminate with the commitment, but reductions are not, so every cleanup done at the wrong time is administrative work with no financial effect.
When should the audience review start?
Ninety days before renewal. That leaves time to reclaim dormant accounts, reassign editions and settle the seat count so the renewal quote is built on the right sized number rather than on last year's, which is the number the quote will otherwise anchor to.
Does edition mix still matter?
Considerably. Between 30 and 50 percent of seats sat on an enterprise edition while consuming only business edition features, and a mixed estate almost always costs less than one company wide edition for the same real capability. Edition mix and seat count are separate levers that compound.
What actually moves a Workspace quote?
Committed seat volume, term length, and the edition mix underneath both. Larger committed counts unlock deeper discount tiers and a multi year term trades for price protection, but each applies to whatever baseline you bring, so the baseline is the first negotiation and the discount is the second.
What evidence should you bring to the renewal?
Active user and feature usage data from the admin console, a leaver and dormant account list, a costed edition mix, and a headcount volatility view that supports your annual versus flexible position. Google negotiates against evidence of real usage rather than against assertions.
Negotiating Google 2: Workspace in Five Minutes
The 300 seat cliff, the exact 20 percent flexible premium, the Vault and storage gates, the January 2025 repricing template, and the renewal true down that is free money.
