Need Help: Who has outsourced or created software programs?

[JDSalinger]

Hello everyone,

This is my first post so first of all thank you for having me here and, I appreciate you taking the time to look at my post.

I have been recently thinking about a few ideas for creating some software programs to later promote to sale through my website. I would not be making it myself and would have someone else do it. It seems the popular method is to use ELance.

I was curious who here has created their own programs or outsourced to have someone else do it?

I have a number of questions, but would like to begin with just hearing about your experiences with it by keying on questions like:

Was it profitable for you?
Was it very problematic or straight forward?
If you outsourced, how to get a good worker?
How did you figure out your budget for this?
How did you go about deciding the program could be profitable?
Deciding on the price etc?
What was your process for addressing questions about the software?
How long did it take for the software to be created? What is a reasonable time frame to expect?
Did you go the affiliate marketing route for your product? If not, why did you choose not to let others promote it?

For a frame of reference, I have been following what Tim Ferriss had mentioned in the 4 hour work week.
For example, he suggests having a 10 times markup on the cost and to shoot for a product selling in the $50 to $200 range.

Looking forward to reading your responses.

[Feenz]

Hi JDS

I am a software developer by trade, and have run software teams.

I've used elance for some web software in the past, but never for installable desktop software.

Unfortunately, its impossible to give too much of an answer to your questions without more detail as to the size of project, what platform its for (eg. Windows), and how specialised it is.

There are some great tools out there for rapid development now, so often things can be achieved very quickly. But my only advice would be to not necessarily go with the cheapest option. Make sure someone is able to very clearly play back to you, exactly what the software is going to do, in their own words, and propose a clear plan and timetable - with penalty clauses. I also would recommend offering a bonus for a specified early finish date - to your approval, and even an additional bonus which is discretionary on your part for keeping in close contact, and your impression of the finished product.

If its not too sensitive for you to tell us what the software is, or if you can think of an alternative product which is similar in complexity, I'd be happy to try to give you a few more pointers.

[JDSalinger]

Hi Feenz,

Thank you for the quick response. The software I am thinking about getting created would be similar to what is available on this site: http://www.ultimatetagranker.com. Avoid the sales hype and just check out video towards the bottom of the page.

My product would be selling to the same audience and have a similar level of complexity.

Let me know what additional information would be helpful. Looking forward to reading your response and thank you for taking the time to help me.

Best,
JDSalinger

[Feenz]

Hi JD

Sorry for not spotting this sooner.

OK, from a quick look, and watching the video on that sales page, it APPEARS to me that this tool interrogates google, and possibly other services, and then provides a report based on those results.

Assuming this is for windows machines only, and assuming I haven't missed something big.

This could be written in a language like VB, or VB .NET.. which is a pretty rapid application development environment, and would use google APIs to interrogate google's results across the internet.

If someone has already built tools using the google API, and therefore would not need time to understand how they work, this could be a very quick piece of softare to write (a really good hard week, or maybe 2 if they encounter some problems).

The things that will give you a much better chance of getting things done quickly, at a reasonable cost, and right are :

1) A REALLY good spec. Spec out every page of the application and how it should look. Spec out whether buttons should be enabled or dissabled, what they should say, what should happen to the rest of the form when you click them. How the results should be gathered, and how they should be displayed (exactly). Try to think of all the anomalies that could happen (what if x didn't work, or what if y returns two things the same, etc, etc).

2) Don't necessarily go with the cheapest. Really think about those who truly understand your project and have spent the time to think about what will be required.

3) Ensure the developer writes a brief spec in their own words as to what the software will do and how it will work, and ideally a top line explanation of how they'll do it - what APIs they'll use, etc. What order they will do what parts in, and how they will structure it.

4) Have set milestones, ask them for help on setting these to suit the project timeline. These milestones should be no more than a couple of days apart, and you need a copy of all code, and to be able to see something working at milestone that are suitable to this.

5) Offer a bonus for completion on time - to your satisfaction, and be clear that this is a bonus. In other words. If you decide you're going to spend $2000 on this project. Try to negotiate the project for $1500, and say you'll give them a $500 bonus for 100%, total completion 24 hour before the due date - and then be really fair on this. You may want to use them again for updates... speaking of which..

6) Look for people who have been around a while, as they are more likely to be around in the future to support it.

7) Finally, make sure you have a copy of the development environment they are using, and ask for the code source to at each stage. Open the project in the dev environment, and if possible, complile and run it. This is absolutely necessary when the project is finished, as you need to be sure someone else could maintain the code if the developer dissapears.

-

I hope this helps. It honestly doesn't look like a massive project, unless I've missed something.

Good luck, and keep us posted.