Over the next 5-6 posts I’ll outline the process which I’ve now followed at several corporates and which I’ve honed to work with my own start-up, Aroxo. I’ll discuss what skills you’ll need, how to write your requirements, how to source developers and designers, how much to budget, how to agree a development contract, how to manage your vendors, how plan your release, all the documentation which you need and much more.

The Aroxo plug

I’ll use Aroxo as my primary reference for how to run the process so it is useful if you know a little about it.

Aroxo is a person-to-person trading exchange for consumer electronics, computer gear, whitegoods, and more. It uses new way to match buyers and sellers that has several benefits over auctions and fixed price sales models. Interestingly it isn’t constrained by the critical mass of other marketplaces – you cannot compete with an existing auction with another auction, no matter what your marketing budget, hence we’ve got a different model which is currently under-wraps.

Marketplaces are inherently complicated systems to build, with lots of workflow which needs building and testing. Especially in our case where we’ve built something which hasn’t been done in the consumer space before. If you’d like to know more may I suggest you sign-up for our beta trial using the sign-up code “bootstrap“.

At 400,00 lines of code, we’ve been up a steep learning curve whilst we’ve bootstrapped Aroxo and we’re passing on the benefits of our learnings here.

What is bootstrapping?

So, what does it mean to bootstrap a company? Bootstrapping involves launching a business on a low budget. Practically this means that you’ll out-source (most likely off-shore) your design and development, you’ll rent your servers, you won’t have an office and you’ll have no salary. Prior to launch, the only expensive professional services which you’ll buy will be your legal advice and accountancy services. Everything else, you’ll have to pick up yourself and learn as you go along.

Why bootstrap? There are a couple of good reasons a company should consider bootstrapping its market entry. The founders may believe that they are onto such a good idea that they don’t want to give up any equity. Or the founders have taken on a small amount of seed financing, just enough to get them into the market. Either way, bootstrapping is a viable model.

There one very bad reason to bootstrap and a major pitfall to be avoided. If no-one would give you any money to build your business, but you’re still convinced you’ve got a great idea then for the sake of my conscience, please think long and hard before spending any money or giving up a source of income. The information I’ll impart here will give you some clear instructions for launching a business, make sure it’s a good way to spend your money before you do.

OK, intros over. On with how to do it.

Overview of the bootstrapping process

I’m starting this series of weekly posts with an overview of the whole process. In the upcoming posts I’ll go into much more detail about each one.

Bootstrapping is a potentially very exciting prospect to an entrepreneur. It enables you get keep a much greater slice of the equity, getting you in to the market so you can raise financing (if you’re going to need it) with a trading history in hand, rather than some PowerPoint. But as I said before, it is fraught with risks and the primary risk which you need to guard against is the failure of the developer to deliver. Two thirds of out-sourced projects fail, whilst this varies from developer to developer, there is a high likelihood of failure – especially when you consider the general start-up failure rate.

Much of the process which I’ll outline is designed to prevent this from happening, so it is worth understanding the main causes:

1. Client fails to precisely express their requirements to the vendor
2. Client continuously changes their mind during the development process
3. A poorly constructed development contract

I could go into detail here with some warning signs and observations about these, but if you faithfully follow the process which I’ll describe here, then you’ve done much to protect yourself.

With this in mind, the bootstrapping process is as follows:

1. Strategy – build your idea into a documented strategy, write down the market you want to serve, what is wrong with how it is currently being served and how you are going to fix it, also think pointedly about how you are going to get the word out to your potential users, where they are online and how you’ll approach them. Test that your idea makes sense
2. Mock-up – build a mock-up of the system you want, each screen in the mock-up must contain the information and navigation you want the final system to contain, you’ll need several goes to get this right, it is not simple
3. Functional Specification – write a document called a functional specification, this explains to the developer what each screen in your mock-up does and what every button and link on the mock-up should do
4. Vendor long-list – build a “long-list” of possible software vendors and NDA them all, but don’t rely on it (more on this later)
5. Request of Information – Write and run an RFI process to sift your long-list of vendors down to a short-list of between 4 – 10 vendors
6. Request for Quotation – Write and run an RFQ process to determine approach, costs, timescales and conditions under which each vendor would deliver, select a preferred vendor and a spare in case of any problems during contract negotiations
7. Award contract – Negotiate a software development contract. There are some essentials to ensure that you’ve got covered, which I’ll go through
8. First trip – if you’ve off-shored then you’ll need to go over there and see the developers to get them started, take them through the entire mock-up and all the documentation, to ensure that they know how the system hangs together. Building a real face-to-face relationship with your developers is not only immensely valuable for you, but also for them too
9. Managing the development – running the development as a client has many responsibilities, you’ll need to know how to deal with delays, systems problems and you’ll need to know how to think like a developer so you can communicate with them. You also need to know when you’ll need to be out there. During this phase you’ll need to be there for the developers to answer all their detailed queries which they’ll have as they start to architect and design the system
10. Alpha, Beta, launch - Aroxo is currently in the process of recruiting beta testers (feel free to sign up using the sign-up code “readwrite”). Getting a large number of triallists and acting on their feedback is essential to building a strong business and having a successful launch, I’ll go through the best way to achieve this

Clearly, whilst I’ve presented this list linearly, it is entirely possible to parallel run some components, e.g. you can start building up a vendor long-list whilst you are doing the mock-up and you can easily be running your RFI and RFQ process when writing your functional specification. But each stage needs to be completed if you are to reasonably expect a quality product by the time the process delivers.

Running through a process such as this is what it means to be an “entrepreneur”, so I’ll be going through each of these stages in plenty of detail in the coming posts.

How much does might this cost?

It is often said that bootstrapping is cheap, and whilst it is certainly cheaper, it is not cheap. Producing a professional and slick product costs money. Our development was a particularly complicated build requiring 5 developers, a systems architect, an html coder, a SQL specialist, 4 testers, 2 designers, a graphic designer, a project manager and a security specialist. If our aim was to raise financing and employ these people directly we’d have burned through $5-7 million before launch. So yes, it is considerably cheaper, but it is not cheap.

With that in mind a rough budget which should be sufficient to bring a system of reasonable complexity to market would be (note that this doesn’t consider any marketing or PR spend):

Project complexity is the main driver of cost, delivery risk and timescale, therefore it is exceptionally hard to be accurate with this, but you are in for a minimum spend of $50k, systems as complex as a trading exchange will involve much bigger numbers. If you have less than this, then you’ll have to do the development yourself, nothing else is going to shift the needle. By doing the development yourself you’re competing with an off-shore company which has specialists in all the skill areas I mentioned above, and there’s a practical limit to the size of the project you can accomplish.

When you consider the question of whether to off-shore your development, if you keep things local, you’ll save on your travel budget, but you development cost will be anything from 5-10 times higher than off-shoring to India or Eastern Europe.

Similarly you’ll need at least one fulltime founder during the development phase to handle the developers and manage any other business development issues which arise during set-up. You’ll notice that there is no budget above for any salary for this individual, so the real cost much higher if you consider lost earnings during the build.

This should give you a good overview of the boot-strapping process, in my next post I’ll outline how to document your requirements in a “functional specification”, how to build a mock-up of the service using free development tools and how to structure your strategy. And I’ll be giving away plenty of examples.

Key takeaways

• If you are out-sourcing design and development, then you’ll want a budget of around USD 50k – 100k, more if you’re developing a complex solution. Anything less and you’ll need to develop the system yourself
• You are not going to be able to manage professional developers part-time
• You absolutely must thoroughly think through how to implement your idea
• Be 100% sure that you have the drive and finances in place to see it through to delivery

Having been through a vendor selection process which encompassed 32 different companies, some off-shore and some on-shore I’ve seen a massive variation in quality, costs and skills. I found the principle of caveat emptor very relevant.

My key learnings were firstly that off-shoring is not cheap, it is cheaper, but it isn’t cheap. Secondly there is a significant variation in quality which a properly run RFI and RFQ process can help select out. Many of the companies we included in our process I would use again, several I would not.

In my experience, there are two pitfalls to avoid when out-sourcing, both of which a professional client can avoid. Firstly the client must prepare a clear specification of what they want – in order to communicate precisely what needs to be developed across timezones and continents the client needs to know exactly what they want. My understanding from talking to off-shore developers is that they frequently receive little information about what needs to be developed.

Secondly, out-sourcing still requires time on the ground with your development team, all the technology in the world won’t replace the accuracy of face to face communication. In addition careful and close management is necessary – speaking to your developer on a daily basis helps to prevent costly errors.

I am also of the view that, assuming the development company has the basic technical skillset to deliver your project then any failure to deliver will be as much in the hands of the client as the developer.

In the end my company we went with a development partner based in India and so far we’ve been very impressed with their professionalism and skills.

See my LinkedIn profile to link to the full range of answers: http://www.linkedin.com/in/mrogers

Yesterday we reached a major milestone and I wanted to share the news with you. We have signed a development contract with an off-shore development company and we are leaving for India tonight to begin the development process. The functional specification is at the printers.

I am personally really excited to be starting the development process and can’t wait to be able to show you more of what we’ve been working on. Our next steps when we’re back we will be to start the patent drafting process, select our site designers and begin work on our detailed marketing plan.

We can’t reveal too much yet as the patent hasn’t been filed, but to give you a little more detail – we’re launching a new online trading exchange, behind it will be an entirely new way of buying and selling – something which has not been done before in the consumer space. You’ll be able to use Aroxo to buy and sell a range of tens of thousands consumer products at extremely competitive prices.

There’s much more to come, watch this space…