The Future of User Behavior – Whiteboard Friday

Hi, Moz fans. I’m Will Critchlow, one of the founders of Distilled, and I want to talk today about the future of user behavior, something that I’ve been talking about a MozCon this year. In particular, I want to talk about the implications of query enhancement. So I’m going to start by telling you what we mean by this phrase.

Old-school query, key phrase, this is what we’ve talked about for a long time. In SEO, something like “London tube stations,” a bunch of words strung together, that’s the entire query, and we would call it a query or a key phrase. But we’ve been defining this what we call the “new query” made up of two parts. The explicit query here in blue is London tube stations, again, in this example, exactly the same. What we’re calling the “implicit query” is essentially all of the other information that the search engine knows about you, and this what they know about you in general, what they know about you at this specific moment in time, and what they know about your recent history and any other factors they want to factor in.

So, in this particular case, I’ve said this is an iPhone user, they’re on the street, they’re in London. You can imagine how this information changes the kind of thing that you might be looking for when you perform a query like this or indeed any other.

This whole model is something that we’ve been kind of building out and thinking about a lot this year. Tom Anthony, one of my colleagues in London, presented this at a conference, and we’ve been working on it together. We came up with this kind of visual representation of what we think is happening over time. As people get used to this behavior, they see it in the search results, and they adapt to the information that they’re receiving back from the search engine.

So old school search results where everybody’s search result was exactly the same, if they performed a particular query, no matter where in the world they were, wherever in the country they were, whatever device they were on, whatever time of day it was, whatever their recent history, everybody’s was the same. In other words, the only information that the search engine is taking into account in this case is the old-style query, the explicit part.

Then, what we’ve seen is that there’s gradually been this implicit query information being added on top. You may not be able to see it from my brilliant hand-drawn diagram here, but my intention is that these blue bars are the same height out to here. So, at this point, there’s all of the explicit query information being passed over. In other words, I’m doing the same kind of search I’ve always done. But Google is taking into account this extra, implicit information about me, what it knows about me, what it knows about my device, what it knows about my history and so forth. Therefore, Google has more information here than they did previously. They can return better results.

That’s kind of what we’ve been talking about for a long time, I think, this evolution of better search results based on the additional information that the search engines have about us. But what we’re starting to see and what we’re certainly predicting is going to become more and more prevalent is that as the implicit information that search engines have grows, and, in particular, as their ability to use that information intelligently improves, then we’re actually going to see users start to give less explicit information over. In other words, they’re going to trust that the search engines are going to pull out the implicit information that they need. So I can do a much shorter, simpler query.

But what you see here is, again, to explain my hand-drawn diagram in case it’s not perfectly beautiful, the blue bars are declining here. In other words, I’m sending less and less explicit information over as time goes along. But actually, the total information that search engines have to work with, as time goes on, is actually increasing, because the implicit information they’re gathering is growing faster than the explicit information is declining.

I can give you a concrete example of this. So I vividly remember giving a talk about keyword research, and it was a few years ago. I was kind of mocking that business owner. We’ve all met these business owners who want to rank for the one-word key phrase. So I want to rank for restaurant or whatever. I say, “This is ridiculous. What in the world can you imagine somebody is possibly looking for when they do a search of ‘restaurant.’ ”

Back then, if you did a search like that, you got a kind of weird mix, because this is back in these days when there essentially no implicit information being taken in. You’ve got a mix of the most powerful websites of actual restaurants anywhere in your country plus some news, like a powerful page on a big domain, those kinds of things. Probably a Wikipedia entry. Why would a business owner want to rank for that stuff? That’s going to convert horribly poorly.

But my mind was changed powerfully when I caught myself. I was in Boston, and I caught myself doing a search for “breakfast.” I went to Google, typed in “breakfast,” hit Search. What was I thinking? What exactly was I hoping the outcome was going to be here? Well, actually, I’ve trained myself to believe that all of this other implicit information is going to be taken into account, and, in fact, it was. So, instead of getting that old-style Wikipedia entry, a news result, a couple of random restaurants from somewhere in the country, I got a local pack, and I got some local Boston news articles on the top 10 places to have breakfast in Boston. It was all customized to my exact location, so I got some stuff that was really near me, and I found a great place to have breakfast just around the corner from the hotel. So that worked.

I’ve actually noticed myself doing this more and more, and I imagine, given obviously the industry I work in, I’m pretty much an early adopter here. But I think we’re going to see all users adopt this style of searching more and more, and it’s really going to change how we as marketers have to think, because it doesn’t mean that you need to go out there and rank for the generic keyword “breakfast.” But it does mean that you need to take into account all of the possible ways that people might be searching for these things and the various different ways that Google might piece together a useful search result when somebody gives them such apparently unhelpful explicit information, in particular, obviously, in this case, local.

I kind of mentioned “not provided” down here. This is my one, I guess, non-
conspiracy theory view of what could be going on with the whole not provided thing, which is that actually, if Google’s model is looking more and more like this and less like this, and, in particular, as we get further over to this end, and of course, you can consider something like Google Now would be the extreme of this where is in fact no blue bar and pure orange, then actually the reliance on keywords goes away. Maybe the not provided thing is actually more of a strategic message for Google, kind of saying, “We’re not necessarily thinking in terms of keywords anymore. We’re thinking in terms of your need at a given moment in time.”

So, anyway, I hope that’s been a useful kind of rapid-fire run through over what I think is going to happen as people get used to the power of query enhancement. I’m Will Critchlow. Until next time, thanks.