Blog: Design Musings and Other Nonsense
We discuss design, business, web products and other miscellany.
No One Wants to Ask Jeeves
You know, I was never interested in asking Jeeves anything. Although it has been several years since I have had a chance (since Ask Jeeves changed their name to Ask.com), this is still something I see today. That is, taking a good idea in the analog world (asking a trusted resource/assistant a question) and shoehorning it into the web. The problem with the approach is, though, that the web is not supposed to be like the real world. It is a different paradigm, with different tools, conventions, and standards. Things work on the web that don’t work in the real world, and vice versa. It has been my experience, at least, that often these types of efforts end up coming off as corny at best and incredibly frustrating at worst.
One contemporary example of this are automated live help agents on websites. The idea, of course, is that human interactive communication can be automated. This is a fine goal, but there is a problem, it doesn’t work. In my experience, and this is only an estimate, live help with a real person is roughly 1,000,000% better than an automated agent. And, why do we need automated live agents anyway? Is live help really that big of a cost sink? Is offering real-live human interaction, something that when people want it has few (if any) substitutes, really that undesirable? Humans are really good at answering contextual questions. And since that is the case, why are we trying to force something that is often vastly inferior?
Now, this may be because a lot of work I do is in User-Experience (UX) design, and I at least try to be considering the end-user as much as I can. As such, it pains me when I see technologies brought to bear that make things worse. Technology is supposed to make things better, and when it doesn’t, something just feels unnatural, like opportunity squandered, like the user is being cheated.
Sure, dealing with humans is sticky. You cannot pre-solve every possible problem. You have to be flexible. Humans are chaotic and you cannot master the chaos with systems. The only way to master the chaos (and turn it into an opportunity), is by empowering (and enabling) people to help other people.
Additionally, there is another problem with using these types of automated systems. They are trying to invent the future by leveraging the past. Sure, a helpful salesperson in a local store is a great thing. And, these types of systems can seem like the obvious solution to offering this type of experience on the web. The problem here, though, is that that logic is flawed. Computers make lousy people, and vice versa. By trying to make computers do a human job, you will probably do it poorly. When instead, you could have used computers for doing what they do best to solve the problem. You cannot solve a problem, especially with something like the web, by looking backward. You are not going to solve the problems of tomorrow with the solutions of yesterday. And, by holding onto the past, you tie an anchor to your efforts.
Sometimes, people solve problems this way because they do not want to have to change how they do things. I get this, I totally do. However, whether they like it or not, the world is changing every day. Not only that, but the changes, especially with the internet, seem to be getting faster and faster. It is almost like we are living in a world where Moore’s Law applies to the complexity of our world.
At the very least, our world is going to get more complex, not less, and the only way to efficiently solve problems is to do so by looking forward. Serve people the way they want to be served, with the best systems capable of doing so. Your users will be glad you did.