Hacker News Reader

Comment by conductr | Hacker News Reader

On the cost side, it’s difficult for me to understand how most startups can pay SV rates for talent and be considered efficient. The problem space usually is just not difficult from a technical aspect, you don’t need top talent. Maybe unpopular here due to the bias of SV talent in HN but from a capital efficiency standpoint startups need to be designed in SV/US and developed elsewhere.

View on HN

We (OP here - Estimote) have a significant team in Europe doing just that. Working on sensor networks. Interestingly our customers are in the US and we went through YC and have mostly Silicon Valley investors.

In one his essays, Paul Graham, pointed out that the number one cost is hiring, so you're not wrong, because hiring bad people can be bad for business, hiring good people can be bad for business if growth stalls.

I guess the key is to hire in proportion to revenue and offer to pay more equity than dollars. This, Paul argues, has employees working harder for the startup, sticking around through tough times, and reducing the burn rate.

The second key thing is, either hire people who can code or can go out and get users.

http://paulgraham.com/startupmistakes.html

The SV salaries are why you're seeing platform B2C companies lead the pack and build full engineering teams in the Bay Area. For many B2B, especially in non-platform markets, building hybrid teams (or building completely outside of SV) with Founders in the Bay and scaled operations often outside the US.

As an example for NFLX with $2.2M revenue per employee ($15.8B rev / 7,100 FTE) if they could save $150,000 per employee it's only $1B to the bottom-line. By hiring "the best" and keeping operations as simple as possible paying more per employee may actually be financially better for them too.

Idk, that “only $1b” is an annual bump to the bottom line. Net income only crossed into the B’s for the first time in 2018. Was significantly less in prior years. The question is how much growth would it take to add another B and how long would it take. At this point, is the market even that big?

Also, Netflix is a great example of a product that could be engineered anywhere with US based design/management (if you’re of the opinion that innovative ideas start here). Especially since they’re on AWS for infra.

Reply on news.ycombinator.com