Microsoft Is Reportedly Cutting Copilot Prices. It’s Also Raising the Commitment.

Microsoft 365 Copilot discounts could be changing…quite drastically.

According to a September 22 report in The Information, Microsoft has told its sales teams they can authorize discounts of 30% to 50% on Microsoft 365 Copilot for enterprises willing to commit to large seat counts and accept additional usage-based charges for premium features. The breakpoints described in the report work out to roughly 30% off above 1,000 seats and up to 50% off above 10,000.

Microsoft hasn’t published anything. It also hasn’t denied the report. We haven’t seen these discounts show up in the field yet, but the timing lines up with two things Microsoft did very publicly this fall: the launch of the revamped Copilot “super app” and a revised pricing model that lowers the per-seat price in exchange for more consumption exposure.

All signs point to Copilot getting cheaper. If it does, the question is cheaper compared to what.

What’s reportedly on the table

Strip away the headline percentage and the deal looks like this: a lower seat price in return for a bigger seat commitment, plus a second cost meter that runs underneath it.

The seat commitment part should be familiar to Microsoft customers. Microsoft wants volume, and the discount tiers are designed to pull you toward a bigger number than you’d pick on your own. If you were planning on 800 seats, a 30% discount that kicks in at 1,000 is a pretty effective nudge.

The second meter is the newer piece, and unlike the discount, it isn’t a rumor. Premium Copilot features now bill through Copilot Credits, and those credits draw down your Azure consumption commitment (your MACC) rather than sitting in a separate bucket. We wrote about why that’s such a significant change a while back. It means a Copilot deal is no longer just an M365 conversation. It can accelerate how fast you burn through an Azure commitment you negotiated for entirely different reasons, or crowd out Azure projects you’d already budgeted for.

Put those two things together and you get a deal where the discounted seat price is the part Microsoft would want you to look at, and the commitment plus consumption is the part that determines what you’d actually pay.

Why now?

If the report is right, Microsoft has a few good business reasons to do this.

Adoption is still a concern over at Microsoft. Plenty of enterprises bought a few hundred Copilot seats as a pilot, watched usage plateau, and never expanded. At $30 a seat with no discount, the business case for rolling it out to 20,000 people was hard to make. A 50% discount at scale would be Microsoft’s answer to that objection.

The competitive picture has also shifted. Google, Salesforce, and a growing list of AI vendors are all fighting for the same enterprise AI budget, and most of them have moved to pricing models that blend seats with consumption. Microsoft trading a fixed seat price for a lower seat price plus usage charges would be Microsoft doing what everyone else is doing, just with more seats to protect.

And then there’s the MACC angle. Tying Copilot Credits to Azure commitments is a clever way to make Copilot usage feel “free” to the business unit turning it on, because it’s coming out of a pool someone else already committed to. It’s also a clever way to get customers to come back for a bigger Azure commitment a year earlier than they planned.

We’ve been saying this to clients for a while now, and we spent most of an August webinar on Microsoft’s AI commercial model on it: the era of the simple discount is ending. Vendors would much rather give you a better unit price in exchange for a bigger commitment and more consumption, because it moves the forecasting risk from their side of the table to yours.

What to do if Copilot is on your renewal

Theoretically, this could be a good deal. If your rollout plan is real and the seats were going to get used anyway, a 30% to 50% discount is money you’d be remiss to leave on the table. The risk is locking in a baseline before adoption has proven itself, then stacking usage charges on top of a seat count you’re committed to.

A few things we’d push on before signing anything:

Make Microsoft put it in writing. Microsoft hasn’t published a program, and your rep may have a different version of it than the one in the press. Get the specific discount, the eligibility rules, the usage terms, and the renewal treatment into the contract itself.

Check how the tiers really work. Do existing seats count toward the threshold? What happens at true-up? Does the discount survive renewal, or does it reset? A 50% discount that evaporates in year two isn’t a 50% discount.

Find out exactly what’s metered. Which features bill in Copilot Credits, at what rate, with what minimums or required pre-purchase, and what does realistic credit burn look like per user? If nobody on the Microsoft side can give you a straight answer, that tells you something. Set spending limits before you turn on anything usage-based. Waiting for the first invoice is an expensive way to learn the burn rate.

Model the MACC hit. Pull up your Azure commitment and figure out what Copilot Credits do to the burn-down. Then decide deliberately whether you want credits funded from MACC or kept separate. Don’t let it happen by default. Our piece on managing Copilot seats and consumption has more on how to set this up.

Negotiate the exits. Ask for ramp schedules, true-down rights, price protection on usage rates, and a path to reduce seats if adoption comes in under plan. Microsoft is asking you to take on forecasting risk, so the terms should give you somewhere to go if the forecast is wrong.

If this guidance sounds familiar, it’s because it is. It’s the same discipline you’d apply to any large Microsoft commitment, and most of it belongs in your renewal prep whether Copilot is in the mix or not.

Where this is ultimately headed

Microsoft isn’t (reportedly) doing anything unusual here. The entire enterprise software market is swapping simple discounts for better pricing tied to bigger commitments and more consumption. That works out fine when the demand is there. When it isn’t, the customer overpays.

If you’re buying or renewing Copilot in the next few quarters, this is a good moment to slow down and do the math. NPI’s Microsoft licensing and cost optimization team can help you validate the thresholds, model the usage and MACC exposure, and negotiate the flexibility you’ll want before you commit.

Buying or renewing Microsoft 365 Copilot? Talk to NPI’s Microsoft licensing experts before you sign.