Hacker News new | ask | show | jobs
by jukkan 253 days ago
8 million active users after almost 2 years from making the $30pupm license available. Less than 2% of Microsoft 365 paying customers choosing to pay extra for Copilot. It's not difficult to see why Microsoft themselves have stopped reporting on AI revenue, as well as not disclosing any official numbers on M365 Copilot sales. Luckily, a source leaked these figures for Ed Zitron to report in his newsletter.
3 comments

Part of the reason is Microsoft's borgy/corpo-confusing service levels.

I have a paid 365 account and couldn't determine from logging in or account info screens if I was on paid or just the freemium version with my 365 plan

In testing out what I did have access to with Copilot, it was incredibly bad compared to ChatGPT or Claude, so I decided not to pay for Copilot whenever I see an ad for it.

Funny how renaming standard Office apps to "Microsoft 365 Copilot" in mobile apps, web home page for office.com etc. did NOT make people realize they should buy an additional $30 plan called "Microsoft 365 Copilot" to actually get to use all features of Microsoft 365 Copilot.

If only MS could have asked Copilot whether their naming strategy makes sense.

The product naming is completely insane. In the span of a year my corporate portal changed name from portal dot office dot com, to Microsoft 365 dot com, and now its copilot something or other dot com.

The portal completely changed in design and the former portal functionality is hidden in a tiny search icon in a knock off chatgpt interface.

The average non-technical corporate user must be so confused.

The confusing nature of product naming is I think pretty typical of MS. It’s just like with personal accounts, where there is no Outlook or Hotmail domain for logging in, it’s still live dot com, and Windows Live hasn’t been a thing for at least 13 years.
And then they have differerent versions of copilot. The consumer one at copilot.microsoft.com, the office one at really hard to remember m365.cloud.microsoft/chat etc. It's really annoying.

And it used to be better. Copilot.microsoft.com would just redirect to the best version you had access to. But they dropped that for whatever reason.

Microsoft did this same thing with Teams and OneDrive back in the day. It took quite a while before they straightened that mess out.

Not to mention ton when they renamed Lync to Skype…

Sharepoint has been a complete mess ever since its inception. It lacked any data organisation features so unless you had really strict processes and oversight it turned into a shitheap within months. This is still really the case.

Even now it's mostly useful as a teams backend. But even that doesn't solve the shitheap problem, as any team you create in teams, or every yammer group automatically creates sharepoint sites which are not very useful as a sharepoint site and clutter up search results.

Users often don't understand that when they share something in teams there is an actual sharepoint behind it and sometimes people fiddle with permissions there and then other people don't understand why they're not allowed to do X or Y because the controls aren't fully exposed in the Teams interface. It's such a mess.

And .NET Core to .NET, not to be confused with .NET Framework
.NET was always a clusterfuck of naming, as in the early 2000s Microsoft was slapping the .NET label everywhere. .NET Passport (now known as Microsoft accounts), .NET My Services, Visual Studio .NET (which was the same as regular VS, just with support for building C# and VB.NET apps added), .NET Server 2003 (Windows Server 2003), et weary cetera.

In this respect, as in many others related to .NET, Microsoft was inspired by Java. In the late 90s, Java wasn't just a language, it was a VM, runtime environment, enterprise platform (Java EE, now Jakarta EE), smartcard technology, remoting protocol, operating system, desktop environment, floor wax, dessert topping...

So Microsoft did with .NET what they now do with Copilot, right? GitHub Copilot (AI-assisted code completion and chat), Microsoft 365 Copilot (the suite), Microsoft Copilot (the chat thingy), Windows Copilot (the chat thingy, but directly on Windows), Copilot for Azure (LLM doing RAG), Copilot for Dynamics 365 (I assume something similar)...
I don't know, does it matter? What I mean is that the product only just, in the grand scheme of things, got rolled out. MS isn't under some instant pressure to make loads of cash off this new feature. Year by year, people will just become used to having a tool, part of their OS, that they can ask questions to to get some general information on whatever topic.

It's easy to pick an arbitrary date and point out the lack of profit. But why don't we pick a date a year from now? Or three years. Or five years.

I think over time users will come to assume that their computing device has, as part of the many tools on it, a tool that they can ask about stuff and get general answers.

Is it profitable right exactly now? No. But I suspect it will become completely commonplace in a few years, and not even worthy of note or comment for the average user.

MS isnt some cash-strapped startup that needs to post instant profits right this moment.

> MS isnt some cash-strapped startup that needs to post instant profits right this moment.

Hmm no, but they are clearly under pressure. Integrating Copilot Chat in office was a pretty big move, and the constant rebranding of everything also feels like they have zero vision and are losing control and scrambling to fix it.

I wonder if Copilot Chat in office will really convince users though. As it doesn't have access to Graph it will lose some of the most valuable 'intelligence'. And thus users will try it, get convinced it's pretty useless and be more hesitant to pay for a license in the future.

I’m pretty sure those figures aren’t global, since very few people at Microsoft have access to that kind of data outside their area of direct accountability.