Is hiring another software house to build my web-based app prototype a good idea?

Find a strong partner with software development expertise. If you can't convince a comptetent IT person to invest their expertise in your site, your idea probably isn't that good. They will be able to get started on the code and add team members as needed. There will be an ongoing combination of employees, consultants, contractors, and other firms during the life cycle of your application.

Imagine driving by a constuction site and pulling one of the carpenters over and asking if he/she can build you a house based on the "blue prints" you drew in your note pad. Do you know as much about building websites as you do about building houses? If you know nothing about home building, hire a contractor.

Create an in-house team to build your product. This team will later support the application and write code for upgrades and revisions.

I haven't heard any outsourced team outperforming a well chosen in-house development team.

Choose the right people and let them build the company's product. As they do, they'll build a good knowledge of the business domain. Nobody will care more than a responsible employee.

Outsourced teams are fast on building your house, but the moment you get the first leak, their responsiveness decreases considerably...

If you hire an outsourced team to build the product and then hire an in-house team you're in double trouble.

Developers have to understand the code and try to fix bugs using it and this will slow down any effort. Programmers usually love to start clean and this is what they'll do at the first chance they'll get.

I would 100% recommend an in-house team, for the following reasons:

1. Consistent Programming Language
2. Familiarity with the code / system
3. You will only need to PM one team (introducing teams to an app / project is not really easy)
4. In house people will (should) have more at stake
5. You will have more control over the code and the entire process
6. An in house team will be more flexible

It is definitely much better to start in-house. If you don't -- plan on re-writing most of the code once the new team comes in place.

First, if your business is the program itself, then you should ideally bring it in-house, because you don't want your entire business to be out of your control. That being said, you may not be able to do this in practice.

The benefit to outsourcing, when done properly, is that you can gain access to a diverse set of skills that you may not be able to hire in-house. As an example, and I run the risk of over-simplifying, you may need a high-end DBA for some work, but not have enough work for that person to justify hiring them full-time. If you outsource the project, that's not your problem.

The problem, however, is in controlling the quality being produced from an outside vendor, which you may be required to support later on. I have had the misfortune to work on a project built this way (as the later support) and we ultimately replaced the entire code-base. Communication with the development team is one of the key issues, and failure to recognize lapses in quality early on can sink an entire project, even when it doesn't have to end up that way.

However, I have worked with clients who for a variety of reasons did not want to build an entire team in-house (price, cash flow, finding a qualified team, risk management). What I advised those clients to do is to find a third-party to manage the team, and outsource the development from that team manager. The third-party (in those cases, my company) acted as quality control over the companies who were actually doing the programming.

In reference to a comment made previously here, an outside team can outperform an internal team if you don't have the ability to hire effectively. This is one of the biggest reasons to have a technical partner - because without the right background, it will be that much harder to manage this kind of risk.

They key word here is "prototype". You may have to throw it away whether you build it entirely in-house, or outsource, or do a combination of the two. But it may take months to build a quality in-house team, whereas is way more important to build a prototype fast.

Also, if you manage to find a good provider interested in a long-term relationship, you may never need to bring the project fully in-house.

At the same time, you'll need someone on staff to help you with the engineering stuff.

So I would start looking for an outsourcing partner and a CTO / VP of Eng.

P.S. It may also be a good idea to outsource testing to a different provider specialized in QA.

If you decide to build this internally, you could take a look at the Agile Platform by OutSystems.

You can download it for free, build your prototype, grow it into your final web application product, and then go from there. You don't need to worry about robustness or scalability, since the platform takes care of that for you (so you just need to focus on implementing the functionality of your product).

Take a look at http://www.outsystems.com/goto/startups for more details on the special conditions for startups.

Hope you find this interesting

Michel

Disclaimer: I work at OutSystems

Outsourcing often means still babysitting.

if you hand it off you still should be able to do it or understand how all the main pieces should work and fit together.

If you're not sure how best it should be put together, get a system/software architect to sit down with you, capture the lighting in the bottle of why you want to do it how you want to do it and how to best build that into a scalable model.

Finding the sweet spot between over-engineering and under-engineering is key in most startups. It's really easy to be lead astray building all sorts of abstraction factories that don't necessarily build value in 1.0

All the best!

In house at start can be more expensive since you have to spend time on finding the right people to work on your project. Outsourcing on other hand can do that for you, but you have to be careful not to fall into a trap of outsourcing: they can show you their best work, but you wont get their best developers, you'll get junior instead.

This is something that many outsourcing companies does.

BUT if you find the right company, you'll get a loyal partner because that company would have interest for your project to succeed since that will bring them ongoing business!

In your position I would recommend to search for smaller outsourcing companies (max 50 employees) because there you can find the best devoted developers with the same cause as you: make your project a success...

Thanks, Ivan

I currently work for a firm that builds software/websites as projects. I am somewhat biased by that.

From personal experience, we have partners for whom we are their development team. I would suggest finding a company to build your prototype rather than building a team of individuals, because of all 6 of the points Jeff Epstein raises above. I reach a different conclusion from Jeff, but all of his points are valid and should be taken into consideration. When hiring a software development partner (notice, partner, not body shop) you should make sure of the following:

1. Get and check references on projects similar to your size
2. The team understands your goals and ideas as you do (they ask enough questions to understand what you envision, not just what you are asking them to build)
3. You are given a schedule of deliverables along with an estimated completion date (don't go with a shop that is going to 'big bang' your project, insist on weekly deliverables so that you can realign as early as possible in the process)
4. Some form of maintenance schedule should be agreed upon from the beginning, along with a warranty period for the code.

Your 'business' will be your application so you will eventually want to have that knowledge and the tough decisions made by you and a competent internal staff.

However, where you are at right now it may be very difficult for you to hire:

• Competent designers, user experience specialists, programmers, system admins etc.

There are a lot of skill sets required to make excellent web applications and rarely can be found by posting up some job ads to hire a developer. The best developers have jobs ;) and if you aren't a developer you don't even have the qualifications to hire one. There are plenty of people who talk a good game.

So using a company initially is probably the best idea, they build many apps all the time, will be in-touch with what the industries are doing, what works.. what doesn't. I would definitely try to find someone local to you that has done this before and has proven results. Don't try to manage an overseas team, you can find local competent people you can meet with and it can be just as cost effective and certainly less stressful and easier to manage.

If the app starts taking off then start to build your internal team.