Ruby is the focus of this blog, but I’ve found myself focusing on business and design oriented issues more lately as the company I’m with moves into a new phase.

This post is more along this lines…

Features are square

This is an idea from one of my co workers. It comes from the urban planning discourse where the problem of designing a city is to make the entire city work together. What often ends up happening, is that various little hamlets develop with nothing in between. You end up with a city full of disparate parts, with nothing keeping them together.

This struck him as remarkably similar to the development of features. You start with a core idea, but you end up adding on many other features. You end up with gaps.

A concrete, but weird, example:

Say we are a team of people who like cats, and we have an application where the core idea is to get people to learn more about cats. We might start out with a well designed cat picker feature, where people could choose the color of the cat, the fur type, and the ear size, and be given a list of cat species that fill their requirements.

You want to charge people 10$/month to use this app, but the customers aren’t paying.

So, over the next few months you add in a bunch of new cat features, where people can chat about cats, and where they can share cat specs with each other and vote on them. However, these features don’t work either. The cat finder feature was never popular enough to drive signups, and most users spend most of their time voting on cats. The question quickly becomes, what’s next? How do we make the cat app pay?

At this point, you can either sit down and think about new features, like adding a cat cam mashup, which would help people select cats by checking out what they look like in various live feeds, or maybe you think it would be good to add a feature that would help people get in touch with cat owners, so that they could connect and buy cats through the site.

Is this a mistake? At this point, you are 3 months into your cat site, but you haven’t even touched your original cat finder code. You have a bunch of well tested, well developed features, but they don’t work together, and your users don’t find any of the features useful enough to pay. Do you owe it to your original vision to try to make the cat finder work? Or do you see it as sunk and decide to move on to the new ideas?

Features are tempting
Features are sort of like a proxy for hope. A new feature is something that the whole team can unite around, and that they can see the possibility of success in. This also has the possibility of being true. But, at least at the moment, my feeling is that filling in the gaps, and refocusing on the core goal that you originally made the app for is the path to paying customers. I wonder if I’m right.