“Outcomes, not features” is Microsoft’s message to Partners for FY27. But if you’re selling Copilot with one message, you’re selling it to nobody in particular. Your customers’ leadership teams are asking four different questions : governance, capacity, cost and adoption and, increasingly, the first reader of your answer isn’t a human at all. It’s their AI agent. This is how to sell the same technology four ways, and why evidence carries weight in every conversation.
I spent a day at a Copilot workshop in London this week, run by Natasha Bradley of Empower M365 and Curo Services, listening to the people rolling this technology out at scale in some of the most complex, regulated organisations in the country. The lesson running through every session was the one that underpins how we work with our Microsoft Partner clients: nobody buys features. People buy outcomes. So when you’re selling Copilot to the C-suite, that means having at least four conversations, each leading with a different outcome.
The CIO and CTO conversation: Copilot governance and data readiness
Two statistics shared at the workshop frame your technical buyer’s world: 71% of leaders cite security and governance as their biggest AI challenge, and only 14% believe they have the right governance structures in place. That gap is where the CIO lives. They’re not asking whether Copilot is clever. They’re asking how to let people use it without losing control.
Reassurance is straightforward once you frame it properly. Copilot honours the permissions and sensitivity labels your customer already has, keeps their data theirs, and doesn’t train on it. Copilot doesn’t create an oversharing problem, it reveals the one you already have. It surfaces everything a user can access, whether or not they should. Years of over-permissive sharing, broken permission inheritance and unlabelled sites sit one prompt away.
That risk is also your opportunity. Getting the data estate in order before rollout — auditing permissions, fixing sharing defaults, building out sensitivity labels — is a valuable piece of work you can lead and charge for. Fix the data and you fix both the risk and the quality of the Copilot outcomes. Positioned well, governance stops being the objection to manage and becomes the reason a nervous leadership team says yes.
The CEO and COO conversation: capacity, and eating the elephant one bite at a time
The chief executive’s question is simpler: what does this give my organisation back? One large NHS body shared the numbers behind its rollout. Drafting the minutes from a three-to-four-hour board meeting used to take three to four weeks — transcription, a draft circulated round senior executives, then board sign-off. With Copilot’s agentic tools, the same job now takes 13 minutes.
That’s the number that sells Copilot. A specific, checkable before-and-after your customer can feel in their own working week — set against Asana’s Anatomy of Work finding that around 60% of people’s time goes on “work about work” rather than skilled work. One NHS lead described their goal in five words: giving employees back their time.
The most useful framing for how to get there came from a South African attendee, who asked me: how do you eat an elephant? One bite at a time. AI adoption works the same way. Find one problem, use Copilot to solve it, prove it worked, then find the next one.
That means your first conversation with a CEO shouldn’t be about Copilot at all. It should be about where their organisation is losing time, where the admin piles up, and where good people are doing work beneath their skill level.
The CFO conversation: Copilot licences, agent credits and proof of value
The finance conversation has changed rapidly, and plenty of pitches haven’t caught up. Part of the problem is that “Copilot” is no longer one thing. It’s an umbrella covering Copilot Chat, Microsoft 365 Copilot, Cowork and, soon, Scout — with different models running underneath. If your own sales team has to pause before explaining the difference, imagine the CFO’s side of the table.
The pricing model has shifted too. Copilot used to be a licence cost. Since Microsoft switched on the meter for Cowork, there are agents, credits and token consumption to think about. An agent pointed at a shared mailbox works around the clock, and every question it answers carries a token cost. That could be a value story or a real cost, and the CFO needs you to present both sides honestly.
The evidence is getting easier to gather. One proof-of-value exercise shared at the workshop, from a property management firm, mapped Copilot Cowork against 28 real business processes: rent arrears, collections, invoice categorisation from a shared mailbox, pulling Yardi data into Excel reports, managing the backlog after annual leave. Each was measured against the average hourly rate of the people doing the work. The proof-of-value group alone evidenced £18,241 in savings, with the annual time saving projected at close to 7,000 hours.
The method matters as much as the number. Deliver Copilot and Cowork in a structured way, mapped against named processes, costed against real salaries. That’s what turns “it feels faster” into a business case a finance director signs against.
The internal audience: Copilot adoption, change fatigue and the six-month test
Beneath your buyers sits the audience that decides whether Copilot actually delivers: managers and end users with day jobs, change fatigue and, often, a healthy scepticism about AI.
The change specialists at the workshop put it well: resistance is signal, not sabotage. It usually points to unanswered questions. Switching Copilot on is nowhere near the finish line. The test that matters is whether people are still using it, and using it better, six months on.
That means your content job doesn’t end at the sale. Prompt guides, use-case libraries, prompt-a-thons, adoption stories, training content for the sceptical middle manager. Support this audience well and your customers renew, expand and refer.
What every Copilot audience has in common: evidence and useful content
Each of the four audiences speaks a different language and needs different content. But all four respond to the same two things: evidence and usefulness.
Case studies work because they carry proof a feature list never can: a real organisation, a real process, a real number. The 13-minute board minutes. The £18,241 across 28 mapped processes. And there’s a newer reason specificity pays: your customer’s own AI agent is increasingly the first reader of your content, summarising you and comparing you against your competitors before a human ever visits your site. Agents can’t cite a vague claim. They can cite a number.
Useful content — the kind that helps a reader do their job better whether or not they ever buy from you — builds the trust that makes the eventual sales conversation shorter.
Copilot doesn’t need another feature list. It needs a better story, told four ways.
Which of your four conversations is weakest right now? That’s where your next piece of content should start.
FAQs
Our Copilot case studies don’t have numbers like £18,241 in them. How do we get evidence that specific?
Build measurement into the delivery, not the write-up. If a proof-of-value maps processes and hourly rates from day one, the case study writes itself. If the project’s already done, interview the people who ran it — a before-and-after like “three weeks to thirteen minutes” usually exists in someone’s head, it just hasn’t been asked for. That’s an interview job, not a marketing one.
Microsoft publishes plenty of Copilot stats and messaging. Why can’t we just use that?
You can, and so can every other Partner — which is exactly the problem. Microsoft’s figures prove the category works; they say nothing about why a customer should choose you. Shared messaging gets you into the conversation. Your own evidence, from your own delivery work, is what wins it.
Should our content admit where Copilot isn’t the right fit?
Yes. Saying “this isn’t ready for clinical decision-making” or “don’t roll this out before your data estate is sorted” reads as expertise, not weakness — to the buyer and, increasingly, to the AI summarising you. Content that only ever says yes gets trusted by no one.
Can Copilot-related content marketing be funded through Microsoft Co-op?
In most cases, yes. Content marketing is a claimable activity under Microsoft’s Co-op programme, and many Partners have ring-fenced funds sitting unspent each half-year. Our guide, The Smartest Way to Invest Your Microsoft Co-op Funds, walks through what’s claimable and how to plan a content programme around it — [link to ebook/Co-op cluster]. If you’re not sure what you’re entitled to, it may already be paid for.
Copilot doesn’t need another feature list. It needs a better story, told four ways. We’re the journalists who find those stories inside your business and shape them for every audience you sell to. Let’s talk.
Information is based on a Copilot workshop run by Natasha Bradley (Empower M365) and Curo Services, London, 13 July 2026.
The Smartest Way to Invest Your Microsoft Co-op Funds
Download our latest ebook: The Smartest Way to Spend your Co-op Funds
Read other articles by Juliet Stott and connect with her on LinkedIn
