First published February 17, 2011 in Mediapost’s Search Insider
Imagine if a search engine was smart enough to be able to anticipate your needs before you know you need them. There it sits, silently monitoring your every move and just when you get a hankering for Thai food (burbling up to the threshold of consciousness), there it is with the hottest Thai restaurants within a 2-mile radius. You didn’t have to do a thing. It was just that smart!
Sound utopian? Then take a moment to think again. Do we really want search to become that smart? Sure, it sounds great in theory, but what would we have to share to allow search to become truly prescient?
The odd thing about humans is that we want our lives to be easier, but we don’t want to sacrifice control in the process. Well, to be more precise, we don’t want to sacrifice control in some situations. It all comes down to our level of engagement with the task at hand and the importance of gut instinct.
Humans have a mental bias towards control. We are most anxious when we have no control over our environment. In fact, even when we have very little control over outcomes (such as in a casino) we fool ourselves into thinking we do. We believe that the way we toss the dice on a craps table (or the hat we’re wearing, or the color of our underwear) has some impact of which numbers come up for us. Factory workers on an assembly line are much happier when they have a button that can stop the line, even if they never use it. We love control and are loath to relinquish it.
Even if a search engine had a 100% success rate in anticipating our intent, chances are we’d feel anxious about surrendering control of our decisions. In fact, this issue has already played out once online. At the height of the dot-com boom, billions of dollars were invested in creating friction-fee online marketplaces. The theory was that certain buying purchases, especially in the B2B marketplace, could be totally automated. In a magazine article for supply chain management in 2000, an industry consultant saw a bright future for e-procurement: “”As long as you understand the business rules for making decisions, there’s no reason why you can’t automate. Why can’t two computer systems – with built in rules – talk to each other?”
It sounds completely rational, but ration has little to do with what we want. We want to feel in control. B2B buying didn’t become automated because we have too much investing in making buying decisions, even when we’re buying widgets for the assembly line, a bank of servers or copy paper in bulk. We don’t trust machines, no matter how smart they are, to make our decisions for us.
What we want is a search engine that guides us, but doesn’t push us. We want a smarter search experience, but we think of it as a filter rather than an arbitrator. Ideally, we want a concierge, who can make informed suggestions that we can then act on.
Could a search engine become smart enough to predict our wants and desires before we’re even aware of them? Possibly, but the other part of that trade-off may be one we’re unwilling to make. How much privacy do we have to give up in order for the engine to know us that well? One of the hottest growth markets is in the area of personal technology. These little bits of tech live with us day in and day out. Consider the Fitbit, a sophisticated motion sensor that tracks our daily movements as long as we keep it with us. This daily diary of our activity (even how restless our sleep is) can be fed directly to the Web. The idea is intriguing, but the reality is a little disconcerting, especially when you think where this technology may go in the future.
As we embed more and more technology into our everyday lives, there is the opportunity to collect signals that could help a search engine (but at this point, the label “search engine” seems wholly inadequate), track behaviors and make very educated guesses about what we might be interested in. Our dreams and desires could potentially be crunched into just another algorithm. Practical? Perhaps. Desirable? I suspect not.
Finally, slumbering just below this discussion is the lurking presence of ultra-targeted advertising, and it’s this that we may find most troubling. If technology someday succeeds in reading our very minds, how can we use that same mind to say no?