Don't optimize at first

There a tendency among technology oriented start-ups to do things the right way. How many of view hear the following:

• We must build infrastructure first.
• We cannot rely on existing products. They won't fit our needs.
• We cannot show clients half-baked products.

Such arguments are neither true nor false. They depend on the situation. If a company is a mature one.and knows its market and how the product serves its clients, then the claim is correct. For example: If the product needs to serve millions of users and there is no infrastructure ready on time, then the company will fail to provide an adequate service. However, if the company is young, it is very dangerous to build an infrastructure or tailor the product to an unknown need. It means losing precious time and gambling on technology.

When I founded my first company more than 10 years ago, my partner an I invented a way to handle an online event where a huge audience can attend. One example is a chat with million of people ( we had a discussion with CNN or having our platform during Larry King). Another example was a auditorium with a teacher and many students capable of asking questions during the lesson (not just a few question). The conceptual problem was how to allow any person to interact without generating so much noise that interferes with anyone. Think of a chat room with 25 people and you see how much noise reside there. Can you imaging what happens when numbers scale up? Well, we found a way.

We were thrilled, our team started build the large-group interaction platform and its GUI. We thought we can isolate the infrastructure and the GUI core components even if the market will change. We planned for the long horizon. Boy, we were so wrong.

• We spend so much time on the wrong things so we had no time to really invest it other features that were basic (i.e. all competitors have them).
• The market changed. Instead of education, we went to online presentations and then to market research. After seeing the real market, the product had to change and also the infrastructure had to adjust.
• The funny thing (or sad thing) was that we could not even reach our technology goals. We spent much precious time and energy realizing that. So instead millions we went in the end for less than a thousand. Turned out that this was OK. No one wanted millions of users.

The lesson is: In the beginning, you can cut corners, use third party solutions, use less than perfect GUI. Don't try to plan the perfect system. Rather, reach your potential users and clients and show them what you have. They love it when you talk to them. If you bring enough value they will also accept a half-baked product. After that, learn what they need and what your product really is. Only then plan carefully your long run.

I know it sounds obvious. Yet, look around and see how many fall into this trap.

Amir