You know that feeling? You talk to your friend about buying something. A few hours later you get an ad on Facebook for that product.
In my case that moment came a few weeks ago, a friend and I were talking about re-selling some baby stuff, and later that day Facebook presented me with advertisements for baby stores.
For starters, it’s probably still a hoax that Facebook is snooping on us through our microphones, and these cases are definitely coincidences. However, I don’t think it would be such a bad idea if some large tech-company were listening in. (Sidenote: I don’t put much trust in Facebook, a little bit more in Google and far more in Apple when it comes to privacy, but that’s another story).
Voice assistants need to listen, not talk.
I don’t mean that large corporations violating our privacy is our future, I just think listening computers are the future of chatbots. Clients often bring chatbots up during our consulting sessions because they have this air of innovation and cutting-edge technology about them. Also, I agree they are PR-friendly, having a full-on conversation with a computer is an excellent idea.
I, however, don’t think chatbots in their current form are useful.
Chatbots are today mostly ineffective. The point of this technology is to help users find information offered by a computer program operating behind a chat interface. The problem, however, is that chat is a very ineffective way to navigate through information. We might use conversations as a way to communicate between humans, but when you apply it to computers, it’s generally clunky, ineffective and frustrating for users.
Well-designed visual interfaces are just easier to navigate and give users far better experiences.
There is a much better use-case for chatbots today: that of the listener, or what I call a “passive chatbot”.
Active chatbots vs. passive chatbots.
A chatbot (or voice assistant) operating in the background of a chat conversation is an excellent use-case for bots. You might be deciding when to meet your friends at your house for dinner, and in your phone automatically offers to place a Deliveroo order for the correct time and place.
Moreover, we can even extend this idea to voice now. Silicon Valley is flooding our houses with smart speakers like the Amazon Echo and the Google Home. These devices are specifically designed to listen to our conversations and react to them.
However, I don’t expect a reaction via audio. I assume there will be a continuous feed of answers on your smartphone. Not with active notifications, but when you open your phone to search for something, you instantly see the queries of the last 30 minutes. Maybe your query was already executed? So search, without search.
A stalking chatbot: scary or not?
Now, of course, this sounds very scary: a company listening in on all your conversations and a service sending you recommendations based on what you say.
Then again most new forms of technology were regarded as scary when they were first introduced. An initial, fearful reaction doesn’t have to limit a certain technology if it builds real value for users.
Evaluating tech-driven innovation.
This goes to the heart of my framework for evaluating technology.
Technology needs to create actual value for users, and that should be the primary consideration in assessing new technologies. Not how much media-attention it receives or how fancy it sounds.
Chatbots generally don’t offer real user value today. A decent visual interface works much better in most cases. A passive, listening bot, on the other hand, builds on our existing ways of interaction and fills real user needs.
Right now I’m building a more formal framework for evaluating new forms of technology-driven innovation. I’m for example mapping all kinds of positive and negative feedback loops that users experience when engaging with new technology. This is currently a work in process, but if you have any thoughts, recommendations or want a sneak preview, feel free to reach out!