Hacker News Reader

Comment by skohan | Hacker News Reader

When I see numbers like this, I wonder what on earth those 55k people will do.

Like I recently interviewed a candidate who was working on a team of over 60 iOS developers working on a single mobile application. I've built fairly complex products with a team of 4-5 people - how would 55 more people manage to keep busy, let alone add value?

I realize Amazon is a huge, multinational corporation with a lot of product offerings, but do they really have enough to work on that they need another small city of tech workers?

Don't get me wrong, being in the industry I'm happy to have big players driving up the price of talent, but part of me wonders if they really understand how scaling works (or doesn't) with tech headcount.

View on HN

Since there's an approximately logarithmic relationship between headcount and productivity, you can keep 60 devs busy doing the work of 5 devs by virtue of each of them becoming as productive as 1/12th of a developer.

That's definitely true of some teams at Amazon. If you're working as an "away team" as they call it, it can take months to get even a one-line change reviewed. It can take a couple of days or more just to figure out how to submit a review request for the "host team".

I'd be curious to hear too!

I knew somebody at Twitter years ago, and he said their mobile teams were bloated and individuals didn't do much. I mean, the Twitter app is quite simple overall, how can you have hundreds working on the UI?

My guess is it becomes hard to contribute quickly at an individual level due to process, lots of overhead with planning, and perhaps codebase not being clean, or just overly large that you don't have good sense of the scope.

Would love to hear from somebody that's worked on a team like this

At large companies, Project Managers and Product Managers are really good at filling up time and creating tasks. They also create processes that slow things down. So instead of making a UI change that a customer requested in 1 day as you can do at a startup, It takes months to get through the design,approval,testing,deployment schedules,etc...

A lot of engineering time goes to building things that never ship. These companies do a ton of A-B testing of small changes in various combinations and only the best performing thing ships. When you have 100M DAUs a 1% bump in engagement (or whatever you're optimizing for) is huge

In my career I definitely moved into smaller organisations. Often they would, over time, get bigger and I would move on.

I probably could have earned more, but working with a few good people and directly impacting the end product made work mostly a pleasure.

If it's something that is multiple countries... I could see 1-2 people per language/region for language/ui/testing.

This was specifically front-end developer headcount. Would you really ever need separate developers for separate localizations?

I guess it would depend on the size/scope of the project. I could see it happening, but also could easily see it not being required for most projects.

What application could you imagine needing a static workforce of 60 developers?

Not a mobile app, but I did see a team of 50 working on one project - enterprise Java in... 2002. It still wasn't all developers - between 30 and 35, then some PM, testers, etc. I can envision projects being that size, I'm not sure I can see a good justification for that state though. I can see 60 devs on a project, just not the need for them, if that makes any sense. I'm accustomed to seeing a lot of people work on stuff getting very little done, mostly talking to each other about how stuff should work, but not doing it.

Hiring 40k would be competitors instantly makes your company/product more competitive.

Reply on news.ycombinator.com