Choosing a niche works. Just choose well.

You may have no doubts about your or your team's ability to build websites, mobile apps, software or web-based services; but you still need to be careful with the biggest uncomfortable truth about the virtual world: that it is full of projects and dreams nobody knows, uses or talk about.

People tend to blame projects failures to a lack of advertising, bad marketing or not having enough time to promote a business. Only few acknowledge that the project was doomed to fail right from the start because it was not well planned.

So here I offer you some advice that may help you avoid seeing how your next project ends up as yet another tombstone of the cemetery of the Web. I hope you will like it.

First, take a look at this simple fact: there are two big kinds of technology projects out there: those that work only virtually (I mean, in computers or mobile phones), and those that connect the virtual world with the real one (like an online shop). What kind of project will you build? That is your starting question.

You also want to remember that there are two main ways to focus a project: It can be either generalist or thematic (also called niche). Experience shows that thematic projects perform better than its generalist peers. Take Yahoo! as a good example. It began as a directory (thematic) and began to expand from there to try to cover mostly everything (generalist). As it did, its popularity declined. When let dropped its directory, what made it unique and famous, it began digging its own grave.

Thematic projects work better in technology because of their simplicity. Do one thing and do it well. Let other projects take care of the other necessities. Adobe does it well with it family of design products.

Yet I am afraid knowing this may not be enough. You need a good niche. And there are other things to take into account to be able to find one.

Your most important need is an audience. Then, you can create a product or service for that audience. Your audience may be broad, like offering films by subscription, medium-sized, like offering a dating service to those who want to meet foreign singles, or small, like when you offer an accounting software only for professionals who deal with tax debt cases. You decide. What matters is that, without a clear audience, your project will have a short expiration date.

The second important thing you need to keep in mind is that technology changes. A decade ago most of the Web was running under PHP and one or more databases. Today most programmers are embracing Astro and reading about Qwik. What will happen in ten years time? Who knows! What you should know is that successful projects adapt to changes.

The third one is that same happens with people. We all get older over time. Your project needs to be ready to welcome new users, but it also needs to be ready to prevent saying good bye to those who outgrew you. Most projects try to achieve this by expanding and supplementing the product. It works, but you need to be careful with that too.

Why? Because, four, you need to expand carefully. Your first users are the core of your business. Any expansion away from the core of your business may motivate your core users to feel between the devil and the deep blue sea, even betrayed at times. I think Google is a good example of this. It began with a search engine, and it expanded to become an advertising company (among many other things, of course). It used to work well as a search engine with ads on certain keywords, but they got greedy. To maintain the success of its advertising business, its owners took the unethical decision to start tracking its users. Millions felt that as a betrayal. They stopped using Google. It is no surprise alternatives appeared short time later. How do they all market themseves? As search engines that do not track.

You do not want that. Respect your core users and you should be fine.

Fifth, before launching a project you must choose a path: to either offer it for free, or to make everybody pay for. My advice is to choose one or the other, and not trying to find a nice place in the middle. Why? Because that does not work. If you offer your product for free, it will be hard to ask your users to pay later. If you offer a minimal or basic plan that works to a point, it is not a secret that people will create several free accounts to avoid paying. Always choose free or paid. Free trials work, though.