Our company is investigating a couple of outsourcing companies and operating models and one that we're starting to lean towards is the Build-Operate-Transfer (BOT) model with a firm that has some specific domain expertise we're looking for.
The model is that:
The advantages for us are lower upfront costs as we build up our operations.
So I'm looking for general feedback from others on their experiences with this sort of model, what to look out for and what are some best practices.
Also I'm interested in any feedback on the financial aspects.
Jason and Shree, thank you for your answers. I'm replying to them here due to the comment length limitation Ultimately our goal is to have an offshore subsidiary as part of the global development team. However in the early stages of the company we don't want the expense and hassle of establishing that subsidiary. We do however want to leverage the cost advantages of an offshore team for a portion of our product. We do only have one product but it's fairly large and we'll keep some of the key development in house but want to outsource other elements.
The concern I have had with traditional outsourcing relationships is that the team members may move around in the company or leave the company, one of the advantages of the BOT model is that we're building a team that will be become employees. So we invest in building a team and eventually take over the operation of that team with this BOT model.
We are also looking at other options including a straight outsourcing model, starting the offshore company now (and looking at headhunting fees) etc but when this BOT model came up I saw that it had some attractive attributes but I didn't have any direct experience with it, hence the question to this community.
I have found that there are multiple firms that offer this model, so it doesn't appear to be that rare. Here's a reference: Build Operate Transfer: Tensilica’s Experience in Offshoring Engineering to India and an update (pdf).
Although I tend to agree with Tim -- that this doesn't sound like a good model in the first place -- I'll try to answer your question.
I don't like the idea of you buying it back from them. I don't see the point in this extra transaction -- they are consultants so why not just pay them as consultants where part of the contract is a "bonus" or something in lieu of purchase? Stock purchases are complex, especially if this is international, so why bother?
I think you need to decide whether these guys are consultants or employees. If they're consultants, they should be paid appropriately and there's no sense in more money later on. If they're employees, why not just use a head-hunter willing to take a delayed fee?
One advantage is "seeding with developers with domain expertise," but this sounds fishy to me. If by "domain" you just mean "familiar with the general technology," this isn't different from any other kind of hiring. If you mean "familiar with our target market," I'm doubtful there's much truth to that.
To answer you last questions about typical models for initial rates and transfers, this isn't typical so I don't think there are normal best practices. Surely there's another, simpler way to manage this.
Finally, to briefly argue my original point, what would prevent another consulting shop -- or them for that matter -- from knocking this off with their own "domain experts?" I can see where this would work -- if the company's success is not dependent on the technology. For example, anyone can knock off a basic Twitter, but it doesn't matter because Twitter has the eyeballs and everyone else has a chicken/egg problem. If you're "in the drivers seat" in that sense, then I think this does make sense, except for the complex transfer stuff.
The conclusion to which I arrived after my 10 yrs in (offshore) outsourcing business is this: outsourcing is like eating fugu - it may be delicious, but you better trust the chef.
So, here are my concerns about BOT model:
This said, about 'delicious' side of outsourcing:
This way you would still be good on cost, would likely get much higher results, and also - not locked down to single service provider.
Outsourcing boils down to a trade of risks and resources, and in the optimum case, all parties win, because
BOT outsourcing is a 'third way' to the classic build-or-buy choice for new operations, and works to the same endpoint - at completion, the operation is yours.
Your question is essentially about how relatively to weight the month-by-month payment for the build-operate phase and the completion payment for the transfer.
The way you've described things, the outsourcing company is going to seed the team. That suggests you will expect a relatively short time before you start to see the results you want. Let's say your endpoint is to get to 100 units of work per quarter by month 24, the profile might be something like this:
Q1=0 Q2=5 Q3=15 Q4=30 Q5=50 Q6=70 Q7=85 Q8=100 Q9=100 Q10=100
If you were doing the build yourself, you would either expect a long run-in - because you'd be feeling your way cautiously - or front-loaded costs - because you'd be investing in premium hires and top-end consultancy support in order to get started well. You have the domain expertise to know what you want by way of outputs and maybe 80% of your existing processes will be transferred. So the final transfer is going to be low risk for you - you will not be dependent on the outsourcer's expertise.
On that basis, I'd be going for the opposite answer than you've suggested - being prepared to pay more in that first year for the heavy lifting, and minimising the transfer price. That still gives you a far better cash flow, and it positions the provider premium against the risk you don't want to manage.
Your way round - concentrating the premium at the end - would better-suited either to a situation where the value doesn't start to flow until towards the end of the project (which is common for a venture involving the first use of new technology), or in the case that the transfer itself is high risk (for instance, when expertise is concentrated in a few individuals).
The two main problems with any offshoring model for startups:
"What are typical models for the transfer price at the end of the operate period? Is it a simple multiple of engineering salaries?" I would ask a valuation expert familiar with this type of scenario.
The "Build-Operate-Transfer" model is valid enough. Several of the outsourcing providers I've worked with in Ukraine state that they're open for this kind of arrangement.
That said, I have never been personally involved in any deals like this. My feeling is that companies who buy outsourced teams want reassurance that they can leave the outsourcing provider, if needed. Thus the providers say "Yeah, we can do that", but they don't really push forward on it -- if possible, they prefer to keep the regular outsourcing agreement going.
IMHO, if you're ready to consider outsourcing (and in general, I think outsourcing is often a bad fit for startups), then it makes sense to pursue this. If nothing else, it could be a way to get better deal terms on team transfer upon cancellation of the agreement. Try very hard to get deal terms which enable you to move your team away.
The default deal terms with most providers are that the team stays with the provider, and you're forbidden to hire any of them for 2 years after ending the outsourcing agreement. This effectively means that if you do great work on building and motivating a great team, then the provider 'controls' the deal, because you can't leave the provider without also losing your people.
Then, about actually operating a subsidiary in an overseas location yourself -- I'm not too sure about that one. I know what it takes to run a company in Ukraine, and the culture and procedures are massively different from my home country -- it's a lot of work to run a company there. Effectively, you will have to hire local management, and you'll be completely dependent on them -- if the overseas management wants to deceive you then they can. Don't underestimate what it requires to run a company in Ukraine, India, or wherever you're looking.