December 12th, 2012
This is a sample of Jared’s keynote presentation from the User Interface 17 conference.
Websites are full of links. How useful these links are in helping users complete tasks is another story. Links have to guide users as they follow the scent of information. A vague or confusing link often leads users down a wrong path and in turn increases their rate of failure.
There are warning signs, however, that users are losing the scent. Using the back button — the “button of doom” as Jared refers to it — and pogo-sticking are indications that your users are having difficulty finding what they are looking for. (As Jared explains, pogo-sticking is when a user moves up and down in the hierarchy of your site, much like being on a pogo stick. )
Ultimately, users are looking for trigger words on your site. If they don’t find the words that will compel them to click, they often turn to search. Jared points out that your search logs are a treasure trove of trigger words. The very links that your users are searching and scanning your site for are what they enter into the search field.
The User Interface 17 conference was a huge success. If you weren’t able to join us this past November in Boston, we actually recorded Tuesday’s entire day of presentations. We’re making this available for OnDemand access. Learn more about this 13-hour audio and video bundle of goodness at UIConf.com.
As always, we love to hear what you’re thinking. Share your thoughts in our comments section.
Jared: The back button predicts failure. When I say predicts failure it means when we see the user hit the back button it means they’ve probably lost the scent. It doesn’t cause failure. Don’t do what the guy did. He went and removed the back button from the browser so people couldn’t hit it anymore thinking that somehow solved this problem and then he pointed his users to me. Don’t do that. It doesn’t work.
The big brother of the back button is pogo sticking. Pogo sticking is what happens when the user jumps up and down in the hierarchy of the site like you’re on a pogo stick. Pogo sticking is this up and down movement. It’s a very Goldilocks type thing. When we hear designers talking about it, developers talk about it, they often talk about it in this very Goldilocks type of way.
Yeah, our users, they’re going to go to this page that has all the links, we call those gallery pages but they say the page with all the links, and they’re going to click on the link that seems the best. When they click on that, they’re going to get to the content but that content’s going to be too hot so they’re going to hit the back button and they’re going to go back to the page with all the links, the gallery page, and they’re going to click on the next best link and they’ll go to that one.
Unfortunately, that one will be too cold so they’ll go back and hopefully that third one will be just right. We have this real Goldilocks thing, too hot, too cold, just right thing going on. That’s how they talk about it, but that’s not how it works. When we look at all the links, all the click streams, and we pull out the ones where there was no pogo sticking whatever, just straight through from start to finish, what we see is 55 percent success rate.
If we look at the links that have pogo sticking in them we see an 11 percent success rate. That’s a dramatic difference. This is this up and down motion through the site.
In fact, on e-commerce sites two-thirds of all purchases happen with no pogo sticking whatsoever. No pogo sticking when they’re making the purchase. Think about that for a second. Everybody who works in e-commerce tells me that customers love to compare their products and they want to go to from one product page to another but that would involve pogo sticking. They’re not doing that.
Instead, what they’re doing is they’re choosing the right product from that list of products, that gallery page. They’re choosing it directly from there and then they’re only going into the product page to make sure they made the right choice. Two-thirds of all purchases happen that way. It’s happening at the gallery page is where the decision is being made, not comparing product pages.
In fact, the more pogo sticking we see in e-commerce the less likely that person is to make a purchase. You don’t want your customers to pogo stick on an e-commerce site. On non e-commerce, the exact same pattern happens. The more we see pogo sticking, the less likely that person is going to actually get to the content they’re seeking on the site.
If your goal is all about getting to the content, you don’t want pogo sticking. The last one of these, the last predictor, has to do with the fact that there are only three ways to get from the home page to that target content page. Either you use whatever search capability is on the site, you go through whatever category navigation you have on the site, or that big ass featured content that the marketing people want there, there’s a chance they’ll click on that.
That’s it. We’ve only got those three ways. That’s our limit. When we look at search, we see a pattern that happens over and over. On every site we’ve ever studied, and we’re talking hundreds of sites, we see the exact same patterns. This is true for every single site except for one site. The way this pattern works is someone comes to this site. Let’s say they’re looking for a license for their dog.
They scan the page to find that license. What happens is the user scans the page for the trigger words, studying every possible place the trigger words could be. If they find the trigger word, they click on it. That’s what people do when they find the trigger word. If they don’t find the trigger word, that’s the case we’re interested in. When they don’t find the trigger word what do they do? They go up to search and type in a keyword.
That’s the pattern we see on every single site, on every page of every site whenever we watch users except for one site. That one site? It happens to be Amazon. On Amazon, when someone comes here instead of scanning the page for keywords their eyes go straight to the search box. Why? Because Amazon has carefully trained every single one of its customers that they never put anything useful on the home page.
Every time you’re sitting in a meeting and someone says we should be like Amazon. Amazon I just go to search. Yeah, we could train everybody that our page is useless. Let’s do that. But that’s not what we do. We put useful stuff on the page so people have to scan it.
What’s interesting are those keywords that they type in. Those aren’t just any keywords. A person scans the page looking for their trigger words. Then they don’t find them, they go up, what do they type into that box? Trigger words. We’ve named this wrong. We shouldn’t call it the search box. We should call it bring your own link. That’s what it is.
These people are creating a link. Every time someone types a search keyword into your page, they’re telling you what link they wanted to find on that page. Which means that your search logs are completely filled with the trigger words that users are trying to find but can’t. You want to go through your search logs and look for those trigger words. In fact, if you can figure out what pages they’re searching from they’re telling you which page you needed to pull the trigger words and put them on right there.
Your users are telling you the stuff all the time. I’m hoping you’re catching this stuff because it’s critical. They’re telling you this.
When we watch people in our click streams use search we see some really interesting patterns. If we pull out click streams where there’s no search at all users succeed about 52 percent of the time. If we pull out all the click streams where users actually did use search, we see success about 30 percent of the time. You’re actually 50 percent more likely to find what you’re looking for if you don’t use the search facility than if you do use the search facility on most sites.
The search facility, which is designed to help you find what you’re looking for, is actually not helping you find what you’re looking for because they’re creating links, crossing their fingers hoping the site brings them to that content, and chances are it’s not matched up very well. You can sit here and you can improve the results of your search or you can take every clue the users are giving you and put those links on the page.
This little company IBM, back when they were in the laptop business, discovered 80 keywords that users kept looking for that weren’t on the pages. They went into their search logs, figured out what pages they were on, put the pages in. This was on the part of the site where people bought laptop computers. Within a month they saw sales increase 300 percent just by adding those trigger words to those pages that people were searching from. 300 percent increase. How’s that for an ROI?
You can sit there and make search better or you can sit there and say we’ve lost scent. That’s what’s going on.