Since 1 August 2026 there has been a proper way for a small business to try Microsoft 365 Copilot before paying for it. Microsoft's Copilot in 30 offer gives organisations with fewer than 300 employees a 25-user, 30-day trial of Microsoft 365 Copilot Business, transacted through a Cloud Solution Provider (CSP) partner rather than a credit card form. The partner sets it up, and Microsoft has built a framework around it: targeting guidance, setup and adoption content, and a 30-day success planner.
Twenty-five seats for a month is enough to get a genuine answer to the only question that matters: would this business pay for Copilot after seeing it on its own work? It is also enough rope to waste completely. A trial where licences get handed out and everyone is told to have a play produces one finding: some people used it, most did not, nobody can say why. This is the plan Frontrow uses to make the 30 days produce a decision instead.
What Copilot in 30 actually is
The mechanics, verified against Microsoft's partner announcement: a new 25-user, 30-day Microsoft 365 Copilot Business trial, built for organisations under 300 employees, available for CSP partners to transact from 1 August 2026. It is partner-led by design. The partner identifies the fit, provisions the trial, and supports the business through the month; Microsoft supplies the framework materials, including a success planner tool released in mid-August.
Two boundaries to be clear on before starting. First, it is a trial of Copilot Business, the SMB Copilot licence, working across Outlook, Word, Excel, PowerPoint and Teams for users on Microsoft 365 business plans. Second, 30 days means 30 days: the clock does not pause for school holidays, end-of-quarter crunch or the two weeks the office manager is away. Pick a month the business will actually be at work.
Pick the 25 users like a study, not a reward
The commonest selection mistake is treating trial seats as recognition: the executive team, the managers, the long-tenured. That sample tells you what Copilot does for people in meetings all day, and nothing about the other 80 per cent of the business. The trial exists to predict what a paid rollout would return, so the 25 should look like the business in miniature:
- Spread across roles: admin and operations, finance, sales or client-facing staff, one or two from the front line of whatever the business does, and yes, a couple of executives.
- Heavy users of the tools Copilot lives in. Someone who writes in Word, models in Excel or triages a full inbox daily will surface value or its absence quickly; someone barely in Microsoft 365 cannot.
- At least three or four sceptics. If only enthusiasts trial it, the results are enthusiasm, not evidence.
- People who will actually show up: agree up front that trial users attend two short check-ins and log what they tried. A seat for a passenger is a wasted data point.
Before any licences go out, do the permissions check. Copilot surfaces whatever each user can already access, and in most small-business tenants that is more than anyone realises. An hour reviewing open SharePoint sites and stale sharing links before day one is the cheapest risk control in the whole exercise.
The week-by-week plan
Week 1: setup and first habits
Provision the 25 seats, run one 45-minute kickoff, and give every user three concrete starters for their role: summarise this week's inbox each morning, draft the client update from the meeting notes, explain what changed in this spreadsheet. Vague permission to explore produces vague results; named tasks produce comparisons. Capture the baseline now, asking each user to estimate time spent weekly on the two or three tasks Copilot will target.
Week 2: role-specific use cases
Push each group one level deeper into their own work: finance recasting monthly reporting commentary, operations turning process notes into checklists, client-facing staff drafting proposals from prior examples. This is the week to consult Frontrow's catalogue of what has stuck in other Australian small businesses, at /insights/copilot-use-cases-that-stick-australian-small-business, rather than inventing from scratch. Hold the first check-in: 20 minutes, each user names one thing that worked and one that did not.
Week 3: measure and prune
By now usage has stratified. Look at who has gone quiet and ask why, because the reasons are the findings: wrong use cases for the role, missing data in the tenant, or genuine lack of fit. Reassign seats from confirmed non-users to fresh candidates if the trial mix was wrong. Keep collecting the numbers: minutes saved on the named tasks, and every instance where output was wrong or needed heavy rework, logged without judgement.
Week 4: the decision pack
Stop introducing new use cases and consolidate. For each role represented: hours saved per week on named tasks, quality incidents, and the user's own answer to "should the business pay for this for you?". Convert to dollars against the licence price and write the one-page recommendation before the trial expires, while access to the evidence is still live.
What to measure through the month
- Active use per user per week, not logins: did Copilot produce something they kept?
- Time on the named tasks versus the week-one baseline, per role.
- Quality incidents: output that was wrong, generic or slower than doing it manually, with the task noted.
- The willingness question, asked twice (day 15 and day 28): would you give up a licence for a colleague who wants one?
The decision at day 30
Three honest outcomes. Buy for a subset: the usual result, licensing the 10 to 15 roles where the numbers cleared the bar, at Copilot Business pricing (US$18 per user per month on the current promotion running to 31 December 2026, or inside the Business Standard and Premium with-Copilot bundles at US$23.50 and US$32). Deploy more broadly: only when most roles cleared the bar, and then via a staged rollout with the governance work done properly, for which Frontrow's structured approach at /insights/microsoft-365-copilot-six-week-pilot-playbook is the natural next step. Walk away: a legitimate outcome that still pays, because the business now knows why, and what would have to change in its data or workflows before revisiting.
Try it
Turn the trial numbers into a business case
Enter the hours saved and headcount from your trial and see the payback maths against current Copilot pricing.
Assumptions
Tune your Copilot business case.
Roles
Live result
$704,668
Net annual benefit
- Active users
- 73
- ROI
- 1788%
- Hours / year
- 8,786
- Payback
- 0.6 mo
- Value saved
- $744,088
- Licence cost
- $39,420
Directional only. Real outcomes depend on licence mix, adoption and which workflows you actually target. Book a review to ground the model against tenant telemetry.
Role-by-role breakdown
| Role | Active | Hours/yr | Value | Licence | Net |
|---|---|---|---|---|---|
| Leadership / Exec | 5 | 920 | $143,000 | $2,700 | $140,300 |
| Managers | 14 | 1,932 | $191,100 | $7,560 | $183,540 |
| Knowledge workers | 42 | 4,830 | $324,187 | $22,680 | $301,507 |
| Sales & client-facing | 12 | 1,104 | $85,800 | $6,480 | $79,320 |
The traps
- Trial is not deployment. Thirty days proves value; it does not do the permissions remediation, labelling and policy work a rollout needs. Budget for that separately rather than assuming the trial did it.
- Permissions hygiene comes first, not after. Copilot in the hands of 25 users is 25 people discovering what the tenant overshares.
- Do not measure the novelty week. Week one usage is curiosity; weeks three and four are the predictor of paid behaviour.
- The 25 licences do not convert by themselves. Someone has to transact the real order before the enthusiasm cools; diarise the decision meeting for day 25, not day 40.
- A trial without a baseline produces vibes. If nobody wrote down the before, the after is unfalsifiable.