Most coverage of new Microsoft features comes out in week one.
That's when everyone has opinions and nobody has real data. The demos look smooth. The use cases sound compelling. And then actual deployment happens, and it's always more complicated than the launch announcement made it look.
Copilot Cowork went generally available on 16 June 2026. Past the launch announcements, there's now enough real deployment experience to have an honest conversation — what it's doing well, where it falls short, and what nobody mentioned at launch that you absolutely need to know before you scale access.
This is not a launch announcement. This is what production actually looks like.
What Cowork Actually Does (Quick Version)
Copilot has always been a prompt-and-response tool. You ask it something, it responds. Useful, but reactive.
Cowork changes the model. Instead of prompting Copilot with a question, you delegate a goal.
'Compile a risk summary from the last three quarterly reports.'
'Review these 40 supplier contracts and flag anything that doesn't match our standard terms.'
'Pull together a briefing note on this client from our SharePoint files and recent email threads.'
Cowork plans the work, executes it across your Microsoft 365 environment, and delivers the output. You don't have to stay at your keyboard while it runs. It runs on GPT-5.5
Thinking and Anthropic's Claude Opus 4.8, with an Auto mode that picks the right model for each task.
Standard Copilot
Copilot Cowork
Opens and reads each report
Extracts and structures the output
You're not there. It still runs.
What Real Deployment Actually Shows
What's Working Well
The tasks where Cowork consistently delivers measurable value share a clear pattern. They're multi-step. They draw on structured, accessible data. And they produce a defined output that previously required a person to spend a significant block of uninterrupted time at a keyboard.
Real examples from the types of workflows where we're seeing genuine results:
• Report compilation from multiple SharePoint documents — what previously took 2-3 hours now runs in 15-20 minutes in the background.
• Review and categorisation workflows across large content sets — contract review, policy compliance checking, document classification.
• Research synthesis from structured internal data — pulling together a briefing from multiple sources without someone having to open each file.
The common factor across all of these isn't the complexity of the task. It's the quality of the data underneath it.
Cowork in a well-organised SharePoint environment, with current and consistently labelled content, produces genuinely useful output. The quality of what comes out is a direct reflection of the quality of what goes in.
Where It's Falling Short
Just as consistent a pattern exists in the tasks where Cowork underdelivers.
These are tasks that require organisational context that doesn't live in documents. Internal relationships. Unwritten process knowledge. Institutional history. The situational judgment that lives in people's heads, not in SharePoint.
Cowork can summarise a document accurately. It cannot tell you which parts of that document matter for this specific stakeholder, or how the CEO is likely to react to this finding, or what the unspoken political context of a situation is.
That's still human work.
The second consistent source of underperformance: disorganised source data. Cowork reflects what it can access. A disorganised, inconsistently labelled SharePoint environment produces disorganised, inconsistently useful output.
Cowork is an amplifier. It amplifies well-organised environments and chaotic ones equally.
The Billing Reality Nobody Mentioned at Launch
This is the part that caught some organisations off guard at the end of July. And it's worth being very direct about it.
Cowork doesn't run on your per-seat Copilot licence alone. It uses Copilot Credits on top of it.
Credits are billed at $0.01 (R0.18 / £0.008) per credit. Every Cowork task consumes credits based on the model used, the volume of context retrieved, the number of tool calls made, and how long the task runs.
To give you a sense of what that looks like in practice:
Light task (e.g. summarising a short document): ~125 credits = $1.25 / R22.78 / £1.00 Moderate task (e.g. synthesising 5 documents): ~500-800 credits = $5-8 / R91-146 / £4-6 Complex task (e.g. full contract review across 20 documents): 2,500+ credits = $25+ / R455+ / £20+
Those numbers are per task. For a team running Cowork heavily, costs can accumulate quickly.
Teams that configured spend alerts and usage caps before expanding access managed their costs without surprises. Teams that enabled Cowork broadly without billing controls had an uncomfortable finance conversation at the end of July.
Configure spend alerts and usage caps in the Microsoft 365 admin centre before you expand Cowork access beyond a small pilot group. This is not optional. It's five minutes of setup that prevents a much longer conversation later.
Two billing options exist: pay-as-you-go at the standard rate, and prepaid annual credits at a lower per-credit rate. Pay-as-you-go is the right starting point for most organisations. Two to three weeks of pilot data tells you what your actual usage pattern looks like. Then you can make an informed decision about whether prepaid credits make economic sense.
Governance: The Step Most Teams Skip
Cowork executes tasks using the permissions of whoever initiates them. It reads email, accesses SharePoint, generates documents, and takes actions in connected apps — all based on whatever access the user already has.
This is why the permissions work that should precede any Copilot deployment matters even more before scaling Cowork. An overshared SharePoint environment doesn't become safer when Cowork starts running in it. It becomes a more efficiently exploited one.
Two things to have in place before you expand access:
• Purview audit logging for Cowork activity — this is included in the GA baseline, but needs to be verified as active. It's what gives you the ability to see what Cowork actually did on behalf of a user.
• Agent 365 governance — if you're using Agent 365, Cowork activity falls under the same governance framework. Agent 365 gives you visibility into Cowork as a managed agent alongside any others in your environment.
How to Start the Right Way
Based on real deployment experience, the approach that produces the best results is consistent:
1. Do a permissions review before your pilot. Cowork's output quality is directly linked to the quality and accessibility of the data it can reach.
2. Pick one specific, recurring workflow for your pilot. Multi-step, structured data, defined output. Not open-ended exploration.
3. Configure spend alerts and usage caps before enabling access. Know what tasks cost before adoption takes off.
4. Define what success looks like before the pilot begins. Run it for two weeks. Measure against your definition.
5. Verify Purview audit logging covers Cowork activity before you expand beyond the pilot group.
The organisations seeing the best results from Cowork are the ones that treated it like any other significant operational change: defined scope, measurement from day one, controlled expansion based on evidence.
Not sure where to start?
GTconsult helps organisations across South Africa, the UK, and the US get more from Microsoft 365 and SharePoint — from custom development and migrations to security, support, and AI solutions.
Keep Reading
Check out our other blogs below:
