By Roger Jack, Elegance Technologies, Inc
If you are starting a Micro-ISV (MISV), then your most important decision is deciding what product to build and sell. I wish that I could give you a secret formula for selecting product ideas that guarantees success, but, sadly, there is no such formula. However, I can give you some ideas that can dramatically improve your odds for success.
Create a stage-gate evaluation process
To start, you need to create some sort of stage-gate process. The idea is that you start with some product ideas and then use objective criteria to gradually pass through various stages-gates until you select a single idea. Robert Cooper developed the stage-gate process for product managers of large companies, but you can use a simplified version for your MISV. We (Elegance Technologies) use four stages and three gates:
|Discovery||Identify product ideas|
|Scoping||Do a simple evaluation to narrow the list|
|Business Case||Create detailed marketing and technical plans for the remaining product ideas|
|Development & Testing||Create the product|
Our New Product Process document shows more details about the process that we use. Please consider this a starting point – you need to account for your unique circumstances. Also, you may want a different process for product ideas that are really easy to implement. If it only takes a week to build your entire product (not just a prototype!), then you may be better off building and showing the product to potential customers. No process can be a complete substitute for good judgment.
The remainder of this article discusses important ideas that you can incorporate into your own stage-gate evaluation process.
Include business fundamentals
Make sure you take into account business fundamentals. The most important business fundamental is that your product idea needs to solve a real world problem. Technical coolness and neat features don’t matter unless they solve a real problem.
The second most important business fundamental is that your product idea should be better than the alternatives. It must be better in at least one way for some group of people. Ideally, it should be dramatically better. Be careful how you define â€œalternativeâ€. For example, our product, Lucid Spec, helps create user interface specifications. It competes not only with other software, but with paper prototyping as well.
Generate lots of ideas
Make sure that you start with a large pool (50+) of product ideas. The quality of your final product selection is directly proportional to your starting number of ideas. If you only have a three ideas, it is easy to convince yourself that one is much better than the other two when, in fact, your â€œbestâ€ idea is only a marginally better than the other two. It is like playing the lottery – the more tickets (ideas) that you have, the better the odds that you will have a winner.
Get real feedback from your potential users
Get feedback from real users at various stages. At an early stage, you can simply write down the idea and show what you have written to users. What you write down should be less than a page. Another option is to create a few static pictures to show to users.
At later stages, you can show a SMALL prototype. Notice that I emphasized the word â€œsmallâ€ because if you are a developer, it is easy to start implementing the entire application before you have finished your evaluation. Don’t do it! The purpose of your prototype is to get additional feedback from your potential customers and clarify requirements. I suggest that you time box your effort to two weeks and create a throw-away prototype.
Make a plan for reaching your customers
You need a plan for getting the word out to your customers. Are you going to use advertising, PR, word-of-mouth, or all three? You need to think about this as part of product idea selection, because it may be easy to reach one type of potential customer than another.
We learned this lesson the hard way. We have a product, C-Sharpener For VB, that automatically converts VB.Net application to C#. It was very easy to reach .NET developers (our customers) through the internet because they tend to visit certain websites. They are also likely to use Google to search for this type of solution. We found that it was much harder to reach potential customers of Lucid Spec, our GUI specification tool. Lucid Spec users are often business analysts or software project/product managers. As a group, there are very few websites where they can be found.
Estimate the development effort
You must try to accurately estimate the development effort because it has a big impact on the ROI and Payback period (see below). Developers are notoriously optimistic when it comes to estimating development time. I know that this is very hard to do, but you have to try. Steve McConnell’s new estimating book might help you.
Create a financial spreadsheet
You need to create spreadsheet that calculates return-on-investment (ROI) and Payback Period. Make sure you create a cost for your time, since it represents an opportunity cost. Also, be careful that your time horizon is reasonable. What is reasonable? In the tech industry, the world can change enough in three years to violate your underlying marketing and technical assumptions. Of course, you may create software in an industry that doesn’t change as fast, so you can adjust your time horizon appropriately.
I know some of you have been reading this and saying to yourself â€œI just want to build my product and skip all this evaluation stuffâ€. Don’t skip it unless you are just working on a hobby rather than an MISV. Whether you agree with stage-gate process above or not, you need to define some sort of evaluation process BEFORE you start evaluating your ideas. If you don’t, it will be like trying to throw darts at a dart board from fifty feet away, blind-folded after a six pack of beer. You may get really, really lucky, but I doubt it.
Roger Jack is President of Elegance Technologies, Inc. Elegance Technologies create products to help improve software development.