Leaning tower of features


One of the most common traps a software company can fall into is ‘feature-itis’, the notion that all your product needs for success is just one more feature.

I know from experience developing very successful products that sometimes even simple software can have significant value for the right customer. At Yes Energy, a company I co-founded, we probably hit our peak ‘$ per customer/per feature’ in year two or three, with relatively simple software, compared to what we shipped in year ten.

After that peak, each new feature had diminishing returns. That doesn’t mean they were bad features, or that we shouldn’t have added them, but for some products we’d doubled or tripled our engineering investment after ten years, without a similar increase in price. But new features can still drive value. Though it’s hard to get old customers to pay more for a product with lots of new features they might not use, those new features could have value to new customers.

But where you can fall into a trap is when a product hasn’t succeeded in the market yet. And I’ve been there before too. It’s easy to start to think that all you need to do is add X, and when X doesn’t work, surely it’s Y. You talk to a few customers and they say they might consider buying if it had Z, another customer wants X, and when you tell them it already has X, comes up with another reason they won’t buy. It’s maddening.

The lack of a specific feature can be a reason why a product hasn’t taken off, or a poor or buggy implementation of a feature can be an impediment to uptake, but we regularly see spectacularly successful tech companies reach large customer bases with minimalistic software. Facebook in 2009 was radically simpler than it is today. Twitter, when it launched, was buggy, crashed a lot, and lacked many of the features that people would come to love in later years, but still it rapidly attained market dominance in its niche.

The lesson to learn here for a new software product is that beyond some level of basic features, your product either occupies a space that has value, what we call ‘product-market fit’, or it doesn’t. The 200th marginal feature isn’t likely to change your product’s position in the marketplace enough to fix a bad fit.

Now, just because you haven’t succeeded yet doesn’t mean you have bad product-market fit, it’s just that whatever your problem is, spending your time and money writing more software isn’t likely the solution. Perhaps you have to spend more on marketing or sales. Maybe you’ve targeted the wrong type of customer. Especially if you are a coder like I am, it can be hard to pull yourself away from the coding solution and apply your efforts in other directions.

This is where I am at with QuirkLocal right now. It’s not perfect, but it’s a good enough implementation of the basic idea that it should sink or swim on its own merits. I am definitely not going to stop developing it, but I need to personally focus more on finding the market and customers for this product, getting the word out, and engaging with like minded communities. All of which is far outside this coder’s comfort zone. It’s intimidating, but also exciting. Something new.