Same Licence, Different Results

Cartoon of an experienced person in a casual wizard hat guiding a colleague who holds a plain stick, with a laptop on the desk

Picture two people in the same team asking Microsoft 365 Copilot to do the same job. Same licence, same tenant, same Monday morning. One ends up with something she can send to a customer. The other ends up with something he quietly deletes.

The tool is identical. Everything around it is not.

Where the difference shows up

The task is simple: a customer has sent a long Outlook thread about a delayed project, and the team needs a recovery update for a Teams meeting that afternoon.

The first person starts by thinking. She knows what the customer is worried about, and which commitments are real and which are wishful. She asks Copilot to summarise the thread, then to separate the decisions from the complaints. She opens the project plan in Planner, checks the dates against what Copilot said, and spots a milestone stated with far too much confidence. She corrects it, then asks for a draft in Word, matching the tone of an earlier update stored in SharePoint.

The second person types “write an update about the delay” and pastes the answer into a document. It reads fluently, is polite, and is wrong in two places. He doesn’t notice, because he never decided what a good answer looked like.

Same wand. The difference was framing the problem, checking the result, and knowing where the real data lives.

It goes beyond prompts

I’m wary of the idea that this is all about prompt wording. Prompts matter, but they’re the visible tip. Underneath sits judgement: knowing which Excel workbook is the trusted one and which is last year’s copy with a different name. Knowing that a SharePoint site full of duplicates gives Copilot a confused view of your business. Knowing when a sensible-sounding answer should make you suspicious rather than relaxed.

None of that arrives with a licence. It comes from experience, a healthy scepticism, and someone being willing to say “that doesn’t look right” out loud.

It changes how I think about buying decisions, too. A business that rolls Copilot out to everyone and calls it done has bought wands. A business that invests in tidy data, clear ownership and shared habits for checking AI output has started building wizards. I’d take the second approach with half the licences over the first with all of them. A skilled person with a plain tool beats a novice with a fancy one most days.

What I’d do on Monday

Pick one real task, such as a weekly report or a customer reply, and have your most careful person and your most enthusiastic person both do it with Copilot. Then compare more than the output. How did each of them get there? What did they ask first? What did they check? What did they ignore?

That comparison tells you more than any vendor demo ever will. It shows you what to teach, what to document and where your data needs attention.

If you buy services from an MSP or adviser, apply the same test to them. Ask less about which products they resell and more about which problems they’ve solved that look like yours. Ask who will actually work with you, and what they did the last time it went sideways.

Tools will keep changing every few months. Framing a problem, questioning an answer and understanding your own information will still be worth having when the next one arrives. That’s what I’d invest in first, and I suspect it quietly decides whether any of this pays off.

Scout, Now Autopilot: How Microsoft’s Always-On Agent Really Works

Cartoon of a woman standing on a stack of documents looking through binoculars and guiding her colleagues forward

Most workplace AI I use still sits there politely, waiting for me to type something. Microsoft Scout is different, and that’s exactly why I think MSPs and SMBs need to understand how it works before someone in the business goes looking for the download link.

Quick housekeeping first. Microsoft introduced Scout at Build in June 2026 as its first “Autopilot” agent. On 25 September it renamed the product itself to Autopilot, describing it as “previously called Scout”. Yes, same name as Windows Autopilot, the device provisioning service, which is entirely unrelated. The docs still say Scout, so I will too.

So what is it, really?

Scout is an agent that works on your behalf without being prompted every time. The current Frontier release is a desktop app for Windows 11 and macOS. You describe a task in a chat, Scout picks the tools it needs, works through the steps visibly, pauses for your approval on sensitive actions, then hands back the result or saves files to your workspace.

The tools are the interesting bit. Scout can read and write files in an approved workspace folder, run shell commands, drive a browser to fill in forms, and connect to Microsoft 365. That covers Outlook mail and calendar, Teams chats, meetings and transcripts, SharePoint Lists, and files in OneDrive, SharePoint and Teams. It creates Word, Excel and PowerPoint files, remembers your preferences and can spin up parallel sub-agents.

What makes it “always-on” are two background modes. Heartbeat runs check-in prompts every 15 to 120 minutes while you’re away, and Automations run on schedules or triggers you define. The September announcement goes further: Autopilot becomes cloud-hosted, with its own identity, memory and workspace inside your tenant, and you @mention it in Teams, Outlook and documents like a colleague.

Under the hood it’s built on open-source OpenClaw, with Microsoft adding Entra identity and access controls on top.

The permissions story is where you need to pay attention

Scout operates with the user’s existing Microsoft 365 permissions. It can’t see anything the user can’t. That sounds reassuring until you remember how many SMB tenants I see with broad sharing links scattered across SharePoint. An agent that tirelessly works through everything a user can reach will find your oversharing faster than any auditor.

There’s a finer detail buried in Microsoft’s own Responsible AI FAQ. In the Frontier desktop app, model interactions are processed through GitHub Copilot, under separate terms. When that happens, some Microsoft 365 protections, such as data residency, retention, sensitivity label enforcement and eDiscovery, don’t apply to that processing. That’s why admins must complete an attestation before anyone can use it. Read that page before you tick the box.

Getting access is deliberately fiddly. You need Frontier turned on in the Microsoft 365 admin centre, an Intune policy on devices, the attestation form, and a GitHub Copilot Business or Enterprise seat for each user. Miss any one of those and sign-in fails without a clear message, which will make for entertaining helpdesk tickets.

How I’d approach it

Treat this as a pilot for a handful of technically capable people, not a rollout. Clean up permissions and sharing first. Keep the approval prompts on and resist the temptation to click “Always allow”. Lock down heartbeat and automation permissions tightly, because those run when nobody is watching. And budget carefully: Microsoft says Autopilot runs on usage-based billing, not the fixed Copilot licence.

I genuinely think agents that follow through are where this is heading. But an assistant that acts is a very different beast from one that answers. The organisations that do well here will be the ones that realised their governance needed to be ready before the agent arrived, not after.

Sources: Microsoft Scout overview (Microsoft Learn), Responsible AI FAQ, Admin access overview, Introducing the new Copilot with Home, Code and Autopilot

Personal AI Speed Isn’t Company Freedom

Cartoon of a founder stepping back while a small team calmly handles office work

Most business owners I speak with already use AI every day. They write faster, research faster, publish more content, and stitch together a few small automations. That is genuinely useful. It is also where a lot of people quietly stop — and that is the problem.

Personal speed feels like progress because your inbox clears and your notes look sharper. The business, though, still waits on the same bottlenecks it waited on last month. The opportunity is not another trick for your own desk. It is building AI into the company so the organisation gains capacity and you get room to breathe.

Look at the whole organisation

I ask leaders to walk the company with a colder eye: Operations, Sales, Marketing, Hiring, Finance, and Leadership. Not “what tool is trendy?” but “where does work repeatedly snag?” Microsoft 365 Copilot only matters when it sits inside those snags — Word, Outlook, Teams, SharePoint, Excel, Power Automate — the surfaces people already live in.

In Operations, that might be the status update that always needs a founder rewrite. Copilot in Teams can turn a noisy channel into a clean summary so the ops lead can act without waiting for your evening catch-up. In Sales, the proposal that sits in draft for a week can start as a Word document grounded in a SharePoint brief, then get shaped in Outlook as a reply that sounds like your firm, not a blank template. Marketing can pull approved language from SharePoint instead of reinventing the story in every campaign. Hiring can turn interview notes from Teams into a structured scorecard in Word before the panel forgets what they heard. Finance can ask Copilot in Excel to surface the variance you used to hunt for on Sunday night. Leadership can stop treating every decision as a meeting when the context already lives with the work.

Where AI should earn its keep

I look for four signals. Where can AI remove a recurring bottleneck — the same chase, the same rewrite, the same “can you just…” every week? Where can it save another hire because one strong person suddenly produces far more without burning out? Where does output still wait on the founder’s typing speed rather than the team’s judgement? And which decisions still depend on you that should not — approvals, prioritisation, a first-cut plan — only because nobody else has a workable draft?

That is where Copilot and Power Automate earn their place. A Power Automate flow can move an approved Word pack into the right SharePoint library and ping the owner in Teams without another follow-up email from you. Outlook Copilot can turn a tangled thread into clear next steps so sales does not need you CC’d on everything. Excel Copilot can draft the commentary Finance would otherwise rewrite after every board pack. None of that is about looking modern. It is about taking load off people who already know the work.

The test I use is simple: if you disappeared for two days, would the company stall for vanity reasons or for structural ones? Personal AI helps you type through the vanity. Company AI attacks the structure.

Freedom is what happens when you are not the bottleneck

Freedom shows up when Operations can close the loop, Sales can send a solid first cut, and Finance can answer a straight question without your calendar. Not because you worked faster at eleven at night, but because the organisation no longer routes every sticky decision through one person.

I am not anti personal AI. Use it. Just do not confuse a faster founder with a freer company. Build AI into the workflows, the libraries, and the routines — then step back far enough to lead.

Prefer Grease Over Gloss

Cartoon of a coach leaving a cluttered workshop with stained sleeves and a sticky-note prototype, colleagues leaning in

I scroll past another perfect feed and feel that familiar sour taste. Neat theories. Spotless creators. Slide decks that look like they have never met a real tenant. I do not learn much from people who only polish. I learn from people who still have grease under their sleeves.

I want the residue of making. Campaigns that actually shipped. Workshops that ran long because someone asked a hard question. Odd little experiments with clients where half the room shrugged and the other half leaned forward. What worked. What failed in spectacular fashion. The soot, not the lab coat. If the work left a mark on your day, I want to hear about that mark.

What I actually look for

In Microsoft 365 land the polished story is endless. Another AI deck. Another thread about how Copilot will change everything if you just adopt the mindset. Fine. I still want the artefacts. Screenshots with the edges showing. Notes scribbled beside a failed run. The version of the story you tell after the meeting, not the one you rehearsed before it.

Show me the Copilot prompt you tried in Word that spat out a policy draft with three confident wrong facts, and how you fixed the grounding. Show me the Outlook draft that saved you twenty minutes on a reply chain until it politely invented a deadline that never existed. Show me the Teams meeting summary that missed the one decision that mattered, so you had to go back into the recording like a detective with a coffee that had gone cold.

I care about the imperfect Power Automate flow that emailed the wrong SharePoint library for a week before someone noticed the condition was upside down. I care about the SharePoint build with three views too many and a permissions model you had to redraw on a whiteboard while the client watched. I care about the workshop demo that half-worked on the day — Copilot Chat answering beautifully until someone asked about their own retention labels and the room went quiet while we dug through the admin centre together.

That quiet is useful. That stumble is useful. The residue of a real session is useful. Perfect creators rarely leave residue. Practitioners leave trails you can follow.

Content theatre versus leftover evidence

Plenty of content is theatre. Lights up. Script tidy. No fingerprints. I am not interested in performing usefulness. I am interested in doing useful work and then talking about it. The difference sounds small until you have sat through both kinds of session and realised which one changed how you configure a tenant the next morning.

The best posts I have written for CIAOPS were not born as content calendars. They came from a messy Tuesday: a Copilot experiment in Word that surprised me, a flow that broke overnight, a Teams rollout where the chat etiquette mattered more than the licence SKU. I made something. Then I showed the leftovers. Colleagues remembered the leftovers longer than the slides.

When I stagger out of a workshop still holding a sticky-note laptop and a prototype that looks like it lost a fight with a highlighter, that is the signal. Not a victory lap. Just: we figured something out, and here is the board if you want to try it yourself. Bring your own stains. Ask awkward questions. Prefer the workshop mess over the feed glow.

In a world of perfect creators and AI slide decks, I would rather be the practitioner with dirty hands. Coach the work. Share the stains. Prefer makers over gloss.

Do something useful and talk about it. Make something and show it. That is enough.

Clear the Runway. Then Cross the Gap.

Warm cartoon office scene with a person at a desk facing a blank document and a colleague cheerfully narrowing a messy whiteboard of sticky notes into three simple cards

I have a habit when someone tells me they are stuck. A client, a mate, a team member staring at a half-finished brief. My first instinct is almost always the same: strip friction out of the process. Make the first step smaller. Hand them a simple framework. Cut a few steps so starting feels less heavy.

That instinct is usually right. Most creative stuckness is external. Too many options. No clear place to put the work. A blank page that asks for a masterpiece before it will accept a rough sentence. In Microsoft 365 we spend a lot of time removing exactly that kind of friction, and we should.

Tools that clear the runway

Copilot in Word is a clean example. The blank document used to be the whole problem. Now you can point Copilot at a previous proposal, a Teams thread, or a SharePoint folder and ask for a first pass. The page is no longer empty. Outlook is similar — a reply draft appears before you have argued with the empty compose window for ten minutes.

Teams and Planner do the framework job. Instead of “we should write something,” you get a channel, a few checklist cards, and an owner. Power Automate cuts the busywork between systems so the human only shows up for the judgement call. SharePoint stops the scavenger hunt for last quarter’s file. External friction falls away. People move.

I have watched this pattern work again and again. Clear the runway and most people take off.

When the gap is inside

Sometimes they don’t. They know exactly what to write. They want to write it. The brief is sitting in SharePoint, the Word outline is already there, Copilot even offered a draft — and still nothing happens. The gap is not between them and the tool. It is between wanting and starting.

That is internal friction, and no amount of Microsoft 365 polish removes it on its own. You can have Copilot ready in Outlook, a Planner board tidy enough to make a project manager smile, and a Power Automate flow that files the finished piece. The document still stays closed.

People invent odd personal tricks for this. A side quest before the real work. A particular playlist. Opening the file and typing one deliberately bad sentence so the page is no longer sacred. Rewriting the same bullet three times in Teams before pasting it into Word. From the outside it looks inefficient. From the inside it is charm and coping mechanism mixed together — a private on-ramp the process diagram never shows.

Make it worth wanting

Here is the part I used to resist. If it is not fun, or at least interesting, it often will not get done. Sometimes the answer is not to make the work easier. Sometimes you make a small part harder on purpose, because that hardness makes you want to engage.

I will pick a fancier Word template than I need. I will draft the hard paragraph by hand first and only then ask Copilot in Word to tidy the rest. I will turn a dull status update into a short story for the Teams channel, even though a bullet list would have been faster. The extra friction is the point. It pulls me over the start line.

So when someone is stuck, I still reach for the external fixes first. Copilot, Planner, Power Automate, a cleaner SharePoint path — clear the runway. Then I ask a quieter question: are they blocked by the process, or by the private gap between wanting and starting? Tools handle the first. The second needs a personal tactic, and occasionally a little deliberate difficulty, so the work feels worth beginning.

Writing Is Still Thinking (Even in the Age of Copilot)

A few weeks ago I caught myself doing something I suspect a lot of people are doing right now.

I had an idea. Instead of opening a blank document, I opened Microsoft 365 Copilot and started asking questions.

Within minutes I had summaries, suggestions, outlines, and options. It was impressive. It was also a little unsettling.

The reason was simple. I had an answer before I had really worked out what I thought about the question.

That’s something I’ve been reflecting on more as AI becomes a bigger part of our daily work. We often talk about AI helping us write, but I think the more important question is whether we’re still using writing to think.

The Value of a Blank Page

For most of my career, writing has been one of the ways I clarified my thinking.

Whether it was a client proposal, a blog post, meeting notes, or an internal strategy document, the process worked because it was slow enough to force decisions. You couldn’t hide gaps in your logic. When you sat staring at an empty page, every weak assumption became obvious.

That’s still true today.

When I open a Word or OneNote document and start typing, I often discover that the idea I thought was clear in my head isn’t quite as solid as I imagined. The act of putting words on a page exposes the fuzzy edges.

The danger with AI is that it can smooth over those edges too quickly. Copilot can generate something that sounds complete before you’ve done the hard work of deciding what you actually believe.

The output may be polished. Your thinking may not be.

Copilot Works Best as a Thinking Partner

This isn’t an argument against Microsoft 365 Copilot. Far from it.

What I’m seeing in organisations is that the most successful users don’t treat Copilot as a replacement for thinking. They use it as a way to challenge and refine their thinking.

A practical example might be drafting a proposal in Word. I often start with rough notes of my own. They are messy and incomplete. Only after I’ve worked through the problem do I ask Copilot to review the draft, identify gaps, suggest alternative approaches, or summarise the key points.

The sequence matters.

Thinking first. AI second.

If you reverse the order, there’s a risk that you’re evaluating someone else’s answer rather than exploring your own understanding.

That’s a subtle difference, but it’s an important one.

Faster Answers Aren’t Always Better Answers

Many of us spend our days moving between Outlook, Teams meetings, client conversations, and project work. The attraction of having instant answers everywhere is obvious.

The challenge is that genuine understanding usually takes longer than information retrieval.

Copilot can tell you what happened in a Teams meeting you missed. It can summarise a lengthy email thread in Outlook. It can pull key points from a SharePoint document in seconds.

That’s incredibly useful.

What it can’t do is decide what those things mean for your business, your clients, or your next decision.

That responsibility still sits with you.

I think that’s why many people who try AI for the first time are both impressed and disappointed. They’re impressed by the speed. They’re disappointed when speed alone doesn’t solve the problem they were really facing.

Knowledge is easy to access. Judgement is still hard work.

The Skill Worth Protecting

The organisations getting the most value from AI aren’t abandoning traditional skills. They’re strengthening them.

Clear writing. Critical thinking. Good questioning. Sound decision-making.

Those capabilities become more important as AI becomes more capable, not less.

The people who ask the best questions get better results from Copilot. The people who understand a problem deeply can spot when an AI-generated answer misses the point. The people who can write clearly can guide AI more effectively.

That’s what I’m watching most closely.

AI is changing how we work, but I don’t believe it changes the need to think carefully. If anything, it raises the value of that skill.

The blank page still matters.

Copilot may help us fill it faster, but the real value comes from understanding what deserves to be written there in the first place.

Copilot in Word

Most people meet Copilot in Word the same way. New blank document, click the icon, type “write me a proposal,” hit Generate.

Out comes 400 words of beige. Technically correct. Completely generic. The kind of thing that could be about your business or anyone else’s.

So they shrug, decide Copilot “isn’t that good,” and go back to writing from scratch.

Here’s the thing. They weren’t using Copilot wrong. They were using the worst part of it and ignoring the best part.

A blank page is the one thing Copilot is genuinely bad at. Give it nothing and it gives you nothing — just confident, well-punctuated filler. The magic isn’t in asking it to invent. It’s in pointing it at what you already have.

What is Copilot in Word, really?

Forget the demos where someone conjures a document from a one-line prompt. That’s the party trick, not the job.

What Copilot in Word actually does well is take your material — a file, an email, a meeting, a folder — and reshape it into something new. The official term Microsoft uses is grounding. The button you press is Reference a file.

That’s the whole ballgame. You’re not asking a stranger to guess what your last client proposal looked like. You’re handing it the last one and saying “another like this.”

So the question stops being “can Copilot write?” and becomes “what do I already have that Copilot can stand on?” For most SMBs, the answer is a lot. Years of proposals, reports, policies, and SOWs sitting in OneDrive, doing nothing.

Step-by-Step: drafting from something you already have

This is the workflow I show every client. It takes about ninety seconds.

Open a new document

Start a blank document in Word, then click the Copilot icon. You’ll get the Draft with Copilot box. Don’t type your prompt yet.

Reference the file first

Click Reference a file (the paperclip), and start typing the filename. Pick the document you want Copilot to learn from — an old proposal, a strong report, a policy you’re proud of. You can attach more than one. Microsoft has the full walkthrough in Draft and add content with Copilot in Word, and the Welcome to Copilot in Word page confirms it only ever looks at the files you choose — nothing else.

Write your prompt

Now tell it what you want, and point it back at what it’s holding:

			
Draft a one-page proposal for Acme Pty Ltd.
Reference: /Northwind-Proposal.docx /Acme-Discovery-Notes.docx
Match the structure and tone of the Northwind proposal.
Keep it under 400 words.

Notice what’s missing? Nowhere do I explain what a proposal is. I’m not teaching Copilot to write. I’m showing it how we write and telling it to do that again.

Keep, regenerate, or bin it

Copilot drops the draft in. Keep it, Regenerate for another go, or Discard. Then you edit like a human — because you still have to.

Why this actually changes behaviour

Before: “Copilot, write me a client report.” → generic mush you rewrite anyway. After: “Copilot, write this report the way I wrote the last three.” → a first draft that already sounds like the business.

That second one is the difference between a toy and a tool.

Here’s the real win for MSPs. Grounding turns your clients’ own document history into an asset they didn’t know they had. Every good proposal becomes a template. Every solid SOP becomes a pattern. The licence they’re already paying for suddenly does something a free chatbot never can — because the free chatbot can’t see their files.

And it’s not just drafting. Flip it around and Copilot becomes a reading tool. Open a long contract, type Summarize this document, and it’ll hand back the gist with citations back to the source — chat with Copilot about your Word document covers that side. Summarize before you read. Draft from what you’ve got. Same tool, both directions.

Don’t ask Copilot to know your business. Show it.

If you’re rolling Copilot out to clients and the first thing you teach them is “type a prompt on a blank page,” you’ve handed them the weakest trick in the box and you’ll get the predictable shrug.

Teach Reference a file first. That’s where the value lives, and it’s the bit nobody stumbles onto by accident.

Reference a file isn’t there to help you write faster. It’s there to make sure what gets written already sounds like you.