Ask YC: How many startups outsource?

There are thousands of extremely talented developers around the world that will work for much less than US salaries. I think every company should at least consider outsourcing as an option, especially if you are working on a "one zebra stripe" product or you are incredibly cash-strapped - which probably describes most young startup entrepreneurs.

That said, I've never outsourced development on anything that is the core technology of a company, be it the product itself or any technology that gives the firm its primary service advantage. I believe that if something can be explained and executed properly by an outsourced team at a significantly lower cost than we can do it ourselves, then it can essentially can be done by anyone else just as easily, so it isn't a worthwhile product for us to pursue. It's just a commodity. It could be an extremely profitable project, of course, but it loses its interest to me. And there are always more interesting projects to work on...projects where I or my team actually have something unique to offer.

I have outsourced several ancillary products to foreign developers, though. Every single one has been a complete nightmare. We used a few major Indian firms, including Wipro, and we even had our own dev leads move there to direct coding. All of the projects went over our time and cost budgets. The solutions were also, for lack of a better word, brittle. The Indian developers were very talented and nice people, but there was just too much communication overhead and something seemed to get lost in the back and forth. A few times, it seemed like we were paying the developers to learn and get up to speed on technology they claimed to be experts in.

We also outsourced a decent-sized project to a team of Ukrainian developers. I was ecstatic to get them (their lead developer had post-doc computer science experience and had worked for the Soviet space program, and he quoted us a salary that was about what we pay our interns. A genius for peanuts.) They finished a very elegant solution for us that was blazing fast and blew away our expectations. It was also finished four months early. But then they held it hostage to get more money. We ultimately negotiated to pay a portion of their "ransom," only to find out they had sold our solution to several other firms for the same price. We had just footed the development bill for our competitors. And good luck getting any kind of legal satisfaction in this type of scenario.

We've outsourced graphic design and the like to freelancers in the US with great success, though. Just budget extra time and money in case they don't deliver.

So, all in all, outsourcing can work out...but it is usually a bigger headache than it is worth and you need to be ultra-cautious.

I've done some outsourcing that worked on a smaller scale than what you described. My rules:

-Only outsource on a project basis.

-Find developers who have done almost exactly what I'm looking for previously, usually via rentacoder.com. Invite them to bid on the project.

-Use escrow.

-Don't outsource to companies, only individuals. Companies will have churn problems and will claim expertise in everything. If you're not doing a bid-based approach, you will get a sales guy to try and talk you up.

I think the majority of the people here creating startups are broke, often in college, and in their early to mid 20s. I'd wager that the vast majority do not have any funding beyond credit cards and family. With that kind of budget, outsourcing isn't even something to be considered.

Beyond that, we're mostly hackers. Why would we outsource what we do best?

We've had success outsourcing some of our non-core activities in our shop (4 p2p hackers working on distributed storage). For example, we've outsourced our website implementation and our Facebook app. Though any one of us could do them, it was much more efficient to draw up a set of screenshots and functional goals along with our API and then get back to working on our core project.

One key success factor is to find a high quality group to work with. We used recommendations, but I've also seen a few 2-person shops use oDesk find people for prototypes of applications they didn't have time/interest to build. The prices are extremely competitive, and often you can hire 3 people and use the best result.

You've got to add value when outsourcing and the project has to be of a particular scale.

A friend recently outsourced development which could be concisely specified. From the UK, he got a student in the US to develop security software for 600 pounds. He installed it on two servers which cost 1000 pounds each. He sold the system to an existing client for 8000 pounds. The student required supervision but it allowed my friend to do on-site consulting while the outsourced software was being developed. The difference in timezone wasn't an issue. Indeed, if you're already working full time then outsourcing to a different timezone can be very beneficial.

However, there isn't much opportunity for a start-up to outsource development. Cost isn't a issue. You should be working in a small field, within your competence. There's also transaction cost which places a minimum size on the project to be outsourced. For a start-up, most projects will be too small to outsource.

If you're thinking of outsourcing a project, think whether you should do it at all.

Let me clear my point. Speaking about outsourcing, I was talking about offshore development, outside US.

Well, still, a lot of startups are always looking for someone, either programmer, or designer, or database developer, etc. There are also a lot of single founders. To start a startup one man is not enough. Even if you are a hacker, it's always nice to have one more hacker.

Here is my point: a lot of startups are looking for someone to join them for a tiny salary plus huge chunk of equity. So why wouldn't they consider outsources, using the same amount of money, they may not give any equity at all or just a small piece of it.

kingnothing: you sound to me like you think that outsourcers require more money, than local people. Did I get you wrong?

mercurio: I'm not talking about corporate level outsourcing, B2B. I'm talking about using power of freelancers. They still stay below the US market rates.

I don't think too many startups outsource. Certainly not the type that YC funds, and certainly not development which is where you would hope to gain an edge.

The issue with outsourcing development is that to judge the quality of the outsourced job and whether it satisfies your requirements, you have to be competent enough to build it yourself. But if you could build it yourself, then you wouldn't be outsourcing.

So if you can, build it yourself, and if not, then hire somebody who has the skills that you lack. You really only should be outsourcing urgent one-off things and highly specialized non-core areas like PR and legal.

I outsourced my email server to google apps.