I build software for a living and recently I have been reflecting on a phrase I often use: customer-centric design.

Why are so many products so frustrating to use? For a lot of them, I feel the answer is that they were not built with customer-centric design. In contrast, I would argue that the tools which are the easiest to use and most useful are.

In a word, customer-centric design means taking into account what the customer wants to do and how they go about doing it. While the idea seems simple enough, it is quite rare in practice.

What makes customer-centric design so rare?

Depending on how deep you want to go, there are a number of factors related to the answer to this question. For now, I want to cover two of them.

decisions are tough

I usually say that C-CD is rare because good design requires decisions. Someone has to decide what to put in and what to leave out and most organizations are surprisingly bad at making decisions.

My usual design example is the humble remote control. My 5 year old Tivo remote is terrific. It fits in my hand and the buttons are clear and easy to operation. Even in the dark, I have no problems using it. In contrast, my brand new PS3 remote is terrible. Even though they have had 5 years or more to perfect it, it still sucks. There are a a lot of tiny buttons that I never use and when I need it, I always press the wrong button. Invariably, I need to turn on the lights an search for the right button.

Tivo understood me as a customer. Sony does not.

Whether the Sony remote was designed by an engineer or it was designed by a committee, I do not know. What i do know is that no one made a decision about the buttons I needed.

If you have ever worked in a major corporation, you already know that few people jump up in a meeting and make pronouncements about how it will be. Most people sit back and wait for someone else to lead or they play it safe with no-answer like "include them all".

"Which of these 45 buttons should we include?" The ultimate decision that doesn't decide anything is "all of them".

To get to Apple's 6-button remote, someone had to decide which 6 buttons were the most important. By choosing 6 buttons and dropping 39, someone is going to complain. Often IN ALL CAPS.

If you choose right, you get a great design. If you choose wrong (or even if you do choose right), you get fired. People do an instinctive cost-benefit analysis and usually opt for the safer course -- no decision. Better to have a mediocre remote with 45-buttons than piss off the CEO's cousin by choosing the wrong set of buttons.

Making decisions is an organizational reason why customer-centric design is rare.

communication and understanding

The other reason C-CD is so rare is communication. I actually wrote this post because I realized that I have been dealing with a perfect example of this situation at work.

In my current position, I use two tools to get things done.

The main tool is built by a team of developers 1,000 miles away. I have never met most of them and we communicate by occasional email or phone calls. They have never actually seen us use the tool or had to deal with the problems we use the tool to solve. While I use the tool every day, they dont know much about what I actually do with it.

As you can guess, using their tool is an exercise in frustration. The information I need to find is often hard to get and a task that could take 3 clicks will usually take 10 clicks or not be doable at all. Even though myself and my team are the customers, the tool was built by developers in ways that made it easiest for the developers. It would be hard for them to do otherwise.

In contrast, I recently built my own tool to support my team with another set of tasks. As a user, I was intimately familiar with the problem domain and the workflow we use to solve the tasks. I use the tool every day and I collect input from my team every day. Every week I set aside some time to take that input and improve the tool. I may be biased, but my tool is a joy to use and after just a few weeks I have seen my productivity increase greatly.

The difference between these two tools is understanding the customer's needs which is essentially a communications problem. Unfortunately it is very difficult to transfer this kind of nuanced information, doubly so if you deal with a consumer product that has millions of external customers.

I am still a proponent of customer-centric design but it is good to recognize that several factors just make it plain hard to do.