How to Choose Computer Software?

My name is Alex. I am CTO (Chief Technical Officer) in MRPeasy. I have about 15 years experience in IT, so I often help choose software to solve a particular problem.

In some areas, I have favourite software; however, in many specific areas, I do not have sufficient knowledge. For example, if a friend is building online store, and wants to add online payment processing, I can propose at least five different suppliers. I have used several of these suppliers in previous projects, but I do not remember the specifics of any business software.

While humanly we would like to feel that we have made the best choice, the truth is that testing all the alternative solutions is not possible. The correct approach is to make quite a good choice within reasonable time.

I propose the following algorithm:

Initially, prepare a long list. With the help of search engines make a large list of potential providers that could tackle the task, e.g., 15-20. If the software leaves a bad impression (for example the website is updated 10 years ago), then do not include this provider in the list.

Next, make a list of 4-8 common formal questions that interest you. For example, whether there are clients in a specific industry. Send your questions to each provider (or its local dealer), then when the responses are received, see how relevant they are. On this basis, select the most suitable providers (5 – 10). This is the so-called short-list.

To those who got shortlisted, you need to send a longer list of questions that should include important functionality you need and important terms of delivery and support, such as price.

Based on the responses received and reliable online reviews, list the software for the testing. First on the list should be a software that is the most likely used candidate for your company.

Next, you need to create test task; it should be a small, simplified problem, which however reflects the specific needs of your company. For example, if there will be hundreds of entries in a real database table, then for the test it is enough to come up with only a couple. However, the task has to do the whole cycle of operations carried out in your company.

Take the first software from the short-list and start the execution of the test task. If the result is successful, this software is your choice. If difficulties persist, it is necessary to seek help or contact support. If the problems are solved, you can move on. If not, take the next software from the list and start to test again.

In summary, the software should be selected based on the performance of thoughtful job, close to reality, and in any case not on the basis of a hasty look at the user interface and thoughtless clicking on different buttons.

The challenge is to narrow down the search area to a reasonable level, to establish priorities, and to stop as soon as a sufficiently good solution is found. The chance that the next decision will be significantly better than the previous is not very big. Thoughtful testing takes time, and it makes no sense to spend time testing more than 1-2 programs.