Full user licences sat where attach or Team Member licensing fit, inflating per user cost by 40 to 60 percent on that population
Dynamics prices three different kinds of user and most estates buy one kind. The gap between what a person does and which licence they hold is where the money is, and it never appears on the quote.
Prepared by Redress Compliance · August 17, 2026 · Microsoft advisory. 15 to 25 Dynamics 365 negotiations advised or benchmarked, 2024 to 2025.
Executive summary
Full user licences were assigned where attach or Team Member licensing fit. Inflating per user cost by 40 to 60 percent on the affected population, and nothing on the quote distinguishes the two.
20 to 30 percent of licensed capability showed no meaningful process usage in the trailing year. Modules renewed on autopilot, because a renewal asks for a module list rather than for evidence that the modules are used.
A 30 percent discount on a bloated map is worse than 20 percent on a clean one. On a map carrying 25 percent waste, 30 percent off the bloated figure costs more than 20 percent off the corrected one.
Benchmark discounts run 15 to 40 percent. Depending on size, growth story, and competitive tension, which is a wide band and the wrong place to start the work.
Three kinds of user, one kind of licence
Dynamics prices access by the depth of what a person does. Most estates assign by the department they sit in, which is a different question.
| Licence type | Who it fits | What goes wrong |
|---|---|---|
| Full user (base) | People who run the process end to end | Assigned to everyone in the department |
| Attach | Full users who also need a second application | A second full licence bought instead |
| Team Member | Light read, approve, and record update roles | Rarely used, because nobody segments the population |
| The module map | Capability actually used in process | Renewed on autopilot, 20 to 30 percent idle |
The arithmetic on the last two rows is worth stating explicitly, because it inverts what a renewal usually optimises for. Take a module map carrying 25 percent waste. A 30 percent discount applied to the bloated map leaves you at 70 percent of an inflated number. A 20 percent discount applied to the corrected map leaves you at 80 percent of a figure that was already a quarter smaller, which is materially less money. The better discount on the worse map loses, and it loses by enough that the sequence is the decision.
The assignment is the price, and nothing on the quote reveals it
Across roughly 15 to 25 Dynamics 365 negotiations advised or benchmarked between 2024 and 2025, the single largest distortion was user type assignment. Full user licences were assigned where attach or Team Member licensing fit, inflating per user cost by 40 to 60 percent on the affected population. Dynamics prices access by the depth of what someone does: a full user runs the process, an attach licence covers a full user who also needs a second application, and Team Member covers light read, approve, and record update roles. Most estates assign by department instead, which is a question about org charts rather than about work.
Nothing on the quote surfaces this. A licence count looks reasonable when it matches the number of people in the function, and the number of people in a function is the figure everyone has to hand. The figure nobody has is how many of those people actually run the process end to end, and producing it requires looking at what users do rather than where they sit. That is why the distortion survives renewal after renewal in estates that are otherwise well managed.
The module map has the same shape. Around 20 to 30 percent of licensed capability showed no meaningful process usage in the trailing year, because modules renew on autopilot. A renewal asks for a module list, the previous list is available, and nothing prompts anyone to test it against usage. Renewing last year's module map without a process usage review locks in that waste for another term, at whatever rate was negotiated.
What follows is a sequencing rule that most renewals get backwards. In roughly two thirds of the Dynamics negotiations benchmarked, restructuring base, attach, and Team Member assignments saved more than the achievable discount improvement, and the two compound when sequenced structure first. Benchmark discounts run 15 to 40 percent depending on size, growth story, and competitive tension, which is a wide band and a tempting place to spend the negotiation. It is the wrong place to start. Map modules to process usage, segment the population by what people actually do, then negotiate the discount last against the cleaned map and the benchmark band. The renewal proposal framework sits in the proposal playbook, the Software Assurance question in the SA brief, and the library in the Microsoft practice.
- Usage exports analysed: inactive accounts, plan right sizing, per user reassignment
- Your renewal quote benchmarked against real closed Microsoft deals
- Every risky clause flagged with the exact quote, the page, and the replacement language
The sequence that works
- Segment the population by what people do, not where they sit, since assignment by department is what puts full licences on Team Member roles.
- Identify the attach cases explicitly, where a full user needs a second application and a second full licence was bought instead.
- Map modules to actual process usage before renewal, because estates routinely carry 20 to 30 percent unused capability into the next term.
- Clean the map before negotiating the rate, as a 30 percent discount on a bloated map is worse than 20 percent on a clean one.
- Negotiate the discount last, against the cleaned map and the benchmark band of 15 to 40 percent.
- Expect the restructure to beat the discount, which it did in roughly two thirds of the negotiations benchmarked, and to compound with it when sequenced correctly.
What the Dynamics negotiations showed, 2024 to 2025
Across roughly 15 to 25 Dynamics 365 negotiations advised or benchmarked:
Per user cost inflation where full user licences were assigned to populations that attach or Team Member licensing would have covered.
Licensed capability showing no meaningful process usage in the trailing year, renewed on autopilot because nothing prompts the check.
In roughly two thirds of the negotiations benchmarked, restructuring base, attach, and Team Member assignments saved more than the achievable discount improvement, and the two compound when sequenced structure first.
Benchmark discounts run 15 to 40 percent depending on size, growth story, and competitive tension. Renewing last year's module map without a process usage review locks in 20 to 30 percent waste at whatever rate is agreed.
Watch the briefing · 4:06Where the Leverage Sits in a Microsoft EAWhy the base the discount applies to decides the bill more than the percentage does.
Your first five moves
- Pull what users actually do, not which department they belong to, and segment full, attach, and Team Member populations from that.
- Reassign the roles that attach or Team Member covers, which is where the 40 to 60 percent inflation sits.
- Test every module against trailing year process usage and drop what shows none.
- Rebuild the map before you discuss rate, since a better discount on a worse map is the losing trade.
- Negotiate the discount last. The Microsoft practice rebuilds the assignment with you.
Frequently asked questions
What is the biggest Dynamics 365 overspend?
User type assignment. Full user licences sat where attach or Team Member licensing fit, inflating per user cost by 40 to 60 percent on the affected population across the negotiations reviewed.
Why does that happen?
Because most estates assign by department rather than by what people do. Headcount in a function is a figure everyone has; how many of those people run the process end to end is a figure nobody has.
What is a Team Member licence for?
Light read, approve, and record update roles. It is rarely used properly because using it requires segmenting the population by activity, which is work nothing in the renewal process asks for.
What is an attach licence?
It covers a full user who also needs a second application. The common error is buying a second full licence instead, which is one of the clearest forms of the 40 to 60 percent inflation.
How much capability sits unused?
20 to 30 percent of licensed capability showed no meaningful process usage in the trailing year. Modules renew on autopilot because the previous list is available and nothing prompts a test against usage.
Is a bigger discount always better?
No. On a map carrying 25 percent waste, 30 percent off the bloated figure leaves you paying more than 20 percent off the corrected one. The better discount on the worse map loses.
What discount is achievable?
15 to 40 percent depending on size, growth story, and competitive tension. It is a wide band and a tempting place to spend the negotiation, which is exactly why it should come last.
Does restructuring beat discounting?
In roughly two thirds of the negotiations benchmarked, yes. Restructuring base, attach, and Team Member assignments saved more than the achievable discount improvement, and the two compound when sequenced structure first.
What should we measure first?
What users actually do. Everything else, the attach cases, the Team Member population, the module map, follows from having activity data instead of an org chart.
When should the discount be negotiated?
Last, against the cleaned map and the benchmark band. Agreeing a rate first fixes the assignment and module errors into the new term at a slightly better price.
