Short and honest: The advertised price of CHF 14.57 per user and month is only part of the calculation. Copilot is an add-on and requires an existing Microsoft 365 base licence. The real Copilot costs (TCO) in the first year consist of four blocks: add-on licence + base licence + one-off rollout and governance + ongoing adoption. Rule of thumb: in the first year, the total cost per user tends to be two to three times the bare add-on price.
The common misconception: "Copilot costs CHF X per user"
When the management team asks what Copilot costs, a single number usually comes up: the add-on price. In Switzerland, Microsoft currently lists a promotional price of CHF 14.57 per user and month (annual subscription, excl. VAT), valid from 1 December 2025 to 30 June 2026 for the first year. After that, the subscription renews at the regular list price of CHF 17.00. If you don't want an annual commitment, you pay CHF 20.40 per user and month (Microsoft Switzerland, 2026).
The problem with that number: it falls short. Microsoft 365 Copilot is explicitly an add-on and is not available on its own. It requires a qualifying base licence, for example Business Standard, Business Premium or Microsoft 365 E3/E5 (Microsoft Learn, 2026). The Copilot surcharge therefore comes on top of the licence costs you already pay, not instead of them. If you currently only have a simpler licence, you may first have to upgrade the base before Copilot can be assigned at all.
One clarification that often gets lost in quotes: the three prices (14.57 / 17.00 / 20.40) belong to the Copilot business surcharge for plans up to 300 users. For larger organisations on an E3/E5 base there is a separate Copilot SKU whose price can differ (Microsoft, 2026). All prices mentioned are per user and month and exclude VAT. For the final Swiss cost, add 8.1 per cent on top.
The four TCO building blocks
If you budget Copilot honestly, you count four cost blocks instead of one. The first is the licence, the other three never show up in the bare add-on price.
| Cost block | What it covers | Order of magnitude (per user/year, excl. VAT) |
|---|---|---|
| 1. Copilot add-on | The Copilot surcharge itself | CHF 175 (promotion) to CHF 204 (list price) |
| 2. Base licence | Prerequisite, e.g. Business Premium CHF 17.80/month | CHF 213 (Business Premium) |
| 3. One-off rollout & governance | Permission clean-up, labels, DLP, pilot project, training | one-off project, not a fixed per-head figure |
| 4. Ongoing adoption | Change support, champion network, training, usage measurement | recurring, depending on maturity |
1. The add-on itself
The familiar item: CHF 14.57 during the promotion, CHF 17.00 regular, CHF 20.40 without annual commitment, per user and month (Microsoft Switzerland, 2026). That is the only number Microsoft communicates openly.
2. The base licence
A typical Business Premium base costs CHF 17.80 per user and month (Swiss reseller prices, 2026). Add the base and Copilot together and you get a good CHF 32 per workplace and month, so the full cost per seat is often around double the bare Copilot price. If you only compare the add-on price, you are leaving out half the bill.
3. One-off rollout and governance
This is the block that surprises most budgets. Copilot accesses all content a person is authorised to see, which makes overly broad sharing in SharePoint and OneDrive, grown over years, visible all at once. In its own guidance, Microsoft describes a multi-stage preparation process as the foundation for a secure rollout: identify sites shared too widely, clean up access, apply sensitivity labels and set up DLP policies (Microsoft Learn, 2026). This clean-up work is not included in the licence price, but it realistically belongs to the rollout costs.
For Switzerland there are benchmark figures. The basic rollout or migration of Microsoft 365 in SMEs with up to 200 employees usually lands between CHF 5,000 and 20,000. For around 50 employees, individual providers quote a range of roughly CHF 5,600 to 14,600 (average approx. CHF 8,900), with consulting hourly rates of CHF 160 to 180. Those are provider benchmarks, not official statistics (Swiss market data, 2025). Specifically for Copilot there is a concrete Swiss entry package: around CHF 9,680 for roughly 5.5 days of assessment, recommendations and an adoption concept (itnetx, Microsoft Marketplace, 2025). That too is a single provider's benchmark, and the effective figure depends heavily on your data hygiene.
4. Ongoing adoption support
A licence alone does not create usage. Independent surveys report that weekly active usage is often only around 20 to 30 per cent of the seats paid for (industry analyses 2025/2026, based among others on Gartner). The consequence: the effective cost per genuinely active person can be several times the list price. Structured change support, communication, a network of key people, training, usage measurement, is regarded as the central lever for value, but it creates ongoing additional effort. This is exactly where it is decided whether the licences lie idle or create value.
How large this block is becomes clear in the Forrester study commissioned by Microsoft (Total Economic Impact, March 2025). For a model organisation with 25,000 employees, around USD 4.4 million was spent over three years on implementation and management plus USD 6.9 million on training and onboarding, compared with USD 5.8 million in pure licence costs. Rollout and enablement together therefore cost more than the licence. Per head that corresponds to roughly USD 1,000 to 1,200 for rollout and training (my own rough calculation from the study figures, spread across 10,000 seats). Important: this is a vendor-funded model calculation for a large organisation, not the actual costs of an SME. In smaller companies the per-head figure tends to be higher.
Sample calculation: 50, 100 and 500 users (CHF/year)
The following table shows the pure licence costs per year, once for the Copilot add-on alone, once for the add-on plus the Business Premium base. All values excl. VAT, excluding the one-off rollout and the ongoing adoption (blocks 3 and 4).
| Users | Copilot add-on only, promotion (CHF 14.57/month) | Add-on only, list price (CHF 17.00/month) | Add-on + Business Premium (CHF 32.37/month) |
|---|---|---|---|
| 50 | approx. CHF 8,742 | approx. CHF 10,200 | approx. CHF 19,400 |
| 100 | approx. CHF 17,484 | approx. CHF 20,400 | approx. CHF 38,800 |
| 500 | approx. CHF 87,420 | approx. CHF 102,000 | approx. CHF 194,200 |
Assumptions: promotional price valid until 30 June 2026, list price thereafter, base licence = Business Premium CHF 17.80/month, all values per year, excl. VAT (Microsoft Switzerland and Swiss reseller prices, 2026). On top of that comes, in every variant, the one-off rollout (Swiss SME benchmark CHF 5,000 to 20,000) plus the ongoing adoption support. International TCO analyses arrive at two to three times the bare add-on price per user in the first year, once base licence, clean-up, training and governance are factored in (provider analyses, 2026), as an order of magnitude, not as a Swiss standard value.
The other side: does it pay off?
As honest as the cost side is, the benefit side is robust if you calculate conservatively. The best-documented figure comes from a real field experiment: in the randomised controlled study "Early Impacts of M365 Copilot" (Microsoft Research, 2025), licence allocation was randomised across 56 companies and more than 6,000 employees. Those who used Copilot at least weekly saved around half an hour per week (31 minutes measured) on reading email alone. Across all licence holders, including those who barely used the tool, it was still around 12 minutes per week on average.
At the optimistic end sits the frequently quoted Forrester figure of around 9 hours of time saved per user and month (TEI, 2025). That one is based on self-assessments from a survey and should be read with caution. Forrester itself deliberately continues on a conservative basis: only 50 per cent of the time saved is counted as productively reused, at a fully loaded cost rate of around USD 38 per hour. You should adopt that 50 per cent correction in any calculation of your own instead of naively multiplying hours by hourly rate.
In Swiss francs, at a fully loaded cost rate of CHF 60 to 90 per hour (my own calculation based on the study figures): conservatively, the measured 30 minutes per week for active users correspond to roughly 2.2 hours per month, so around CHF 130 to 195 of value per person and month. If you are extremely cautious and take only the 12-minute average, you end up at around CHF 50 to 80.
What matters is the point from which it pays off. The Copilot surcharge of CHF 14.57 per month is already covered if each person saves about 10 to 15 minutes per month, so only around 2 to 4 minutes per week (CHF 14.57 divided by CHF 60 to 90 per hour). Even the most conservative measured value from the field experiment, 12 minutes per week averaged across all licence holders, is roughly three to four times above that threshold. Put differently: for the add-on not to be worthwhile, employees would have to derive practically no benefit at all. If you include the base licence, the threshold rises, but it stays in the range of a few minutes per week.
You don't have to take my word for this calculation, check it with your own numbers. In the interactive Copilot ROI calculator you can work through your own case, with your hourly rate, your number of users and a realistic usage rate.
What to do now
The ROI does not arrive with the licence purchase, it arrives with the rollout. The Forrester model calculation projects a three-year ROI of 132 to 353 per cent for SMEs, but only with a genuine change programme (Forrester/Microsoft, 2024). Consulting sources additionally cite payback periods of around 5 to 7 months with a programme, compared with 12 to 18 months or no return at all without one (provider analyses, 2026). That is the real argument: support instead of a mere licence purchase.
In concrete terms, this is what it means for the management team:
- Budget the full cost, not the add-on price. Plan for all four blocks, add-on, base licence, one-off rollout, ongoing adoption.
- Clean up first. Permissions, overly broad sharing, labels and DLP belong before the rollout, not after it.
- Run adoption as a project. Key people, training and usage measurement determine the usage rate, and with it the ROI.
- Start with a pilot project and measure the actual time saved before you roll out broadly.
I don't sell licences, I help you get value out of the ones you already pay for. If you want to know what Copilot realistically costs and delivers in your organisation, take a look at my services or book a free initial consultation. And if your licences are already lying around unused today: the TCO only adds up with adoption, and you can read why licences lie idle in the sister article.
