Hacker News Reader

Comment by helsinkiandrew | Hacker News Reader

I'm realizing more and more that relying on a large organization for a service where you are 1 of hundreds of millions of customers is a bad idea.

First customer service will be automated or even non existant, and very poor. Secondly the product will have been 'tweaked' so many times for new markets and product extensions that it will be very fragile when you do something at the edges of its functionality (not what the other hundreds of millions are doing).

It shouldn't really be this way - it tells a lot about software engineering that a product run by a few enthused people alone can often (but by no means must) have better support and service than a product with huge resources.

View on HN

A father with a daughter that goes to a school, and the father has a Microsoft account, is nowhere near an "edge of functionality". Sounds more like Microsoft has a too-complex login system on their hands, one that needs a ton of backend rework to happen in order to simplify the system for the user (and the developer)'s sanity.

Also, a software engineering product run by a large organization is going to have tons more functionality under its much bigger umbrella compared to a small team with a much smaller product. Consider AWS vs Digital Ocean. Both great companies, but AWS's umbrella of offerings is vast compare to Digital Ocean.

> is nowhere near an "edge of functionality"

No, but something unusual went wrong and getting it fixed will be harder at MS than a smaller company. There probably isn't a single person who understands why without a fair amount of research. Without the publicity, MS would be inclined not to spend the effort to fix.

> Also, a software engineering product run by a large organization is going to have tons more functionality

This was exactly my point - the large product with tons more functionality will likely be more brittle, harder to use, and get support for if something breaks. If you aren't using that functionality, you often won't be well served by the company. I had this experience with EverNote. I'm also a very happy AWS customer, but I think that is because their products are a set of (fairly) independent products, rather than one huge system.

Ah. It's because it's nowhere near the edge of functionality, I don't think something particularly unusual actually went wrong here. Mr Jeff just hit a bug. But because Microsoft is such a huge company, and fathers with daughters who go to a school is such a large segment of users (and because Microsoft is not as obtuse as Apple, where Apple pretends their software has no bugs), I bet there's already a product manager spinning up a customer service bulletin going out to reps to school IT departments and a team of developers to add the fix to a roadmap and fix the problem globally.

Or maybe the bug's just going to languish and our friend Jeff here is going to be forced to create a new Microsoft account.

As far as AWS' products being independent products. I have the opposite view, as their products would be rather useless if they didn't interoperate with each other. How useless S3 would be if it couldn't talk to anything else!

The problem is not that there is no one person it's that the team that knows the details are insulated from end users by layers of layers of bureaucracy needed to keep them from being flooded in identical/trivial tickets to the point where they never get to look at the actual product they are developing.

With smaller outfit's the L2 tech support might actually be having a line of communication directly to engineering, where in large companies there might be an L3-6 plus different product owners and escalation managers involved before a case gets in front of the engineering team.

Even in large companies it depends a lot on the product. For Microsoft, if something has an issue tracker on GitHub (e.g. VSCode, C#, Windows Terminal), the bugs that get filed there are usually looked at directly by the team that owns it.

Reply on news.ycombinator.com