Normally, a 404 Not Found error is shown to visitors when they try to visit a non-existent page. But what about when there is a page there, only it doesn’t have what they want – what do you do then? One solution is to offer them a link to the right page, duh! Sounds simple, but it can actually be a bit tricky. Another is to update the page and answer people’s question.
Hattip to Steve Krug, usability maven behind the book Don’t Make Me Think, for inspiring this idea. Hattip to Steve Krug for having the idea first, sigh. All I’ve done is come up with an [slick] implementation…
What I was going to write was based on a navigation example Steve gave where a store might have these three categories: “Power Tools,” “Hand Tools” and “Sanding and Grinding Tools.” Wouldn’t some of the Sanding and Grinding Tools be properly found in Power Tools? And vice versa?
Some people like me would hesitate to click anything before knowing for sure which was the better choice! And they might leave if the first link was unsatisfactory.
So what I was thinking is that you could create a script that would show a help message that would activate depending on user behaviour.
Here’s the help message, and I’ll explain when it would activate in a moment.
“Having trouble finding an item? Some items in this category as well as related items are also categorized as “ABC Cat” (linked) and “DEF Cat,” which might have what you’re looking for.”
For example, someone visits your store and goes to a category page. They browse around for 20 seconds and click nothing. Your script notices this, and then makes the above phrase appear, with the related categories. Another behaviour might be what Avinash calls “pogo-sticking,” where a visitor goes between a category page and several detail pages without adding anything to cart. It often indicates that they’re not finding what they wanted.
On page 157 of Don’t Make Me Think’s Second Edition, Steve has exactly this kind of situation in mind in a footnote. He writes:
“You may be thinking, Well, why not just put it in both categories? In general, I think it’s best for things to ‘live’ in only one place in a hierarchy, with a prominent “see also” crosslink in any other places where people are likely to look for them.”
While I’ve advocated putting things in only a single category before, for the sake of better indexing, it’s something I’m seriously rethinking. To me, from a usability perspective, it makes more sense to help someone find something right away, by putting it everywhere they might look.
Steve’s point is that them trying, something, not finding it, then going back and getting it right on try 2 is just fine as well. I guess that works for some sites (more news/e-commerce sites, I would guess), but I think most blogs won’t necessarily have it that way where the second click is any better than the first.
Anyways, perhaps you might still make better use of screen real estate by making the links and phrase’s appearance conditional on user behaviour that indicates confusion.
Does anyone have a clue how to code this without tripping search engine filters for hidden content?
Another solution to this is to look at the intent of visitors coming to that page, by comparison to competitors’ interpretations of that intent. As Steve describes in this interview with Marketing Sherpa:
“SK: I haven’t done enough with analytics. I’ve read a lot but I’ve never really done enough with them to have a sense of what’s most useful with one exception which is that I always tell people you really wanna look at your search logs. The search logs are enormously valuable and mostly neglected. So, people should look at their search logs.
I tell them you should have a routine where, once a month, you look at the search logs and you look at the top five or ten things in the search logs and execute that search yourself and see what you get in the search engine. You can usually figure out what people are looking for from those top five or ten searches.
“You wanna make sure that when you execute that search, the first search that shows up in the search results is the best page that you have that meets that query. And if you have to rig the results, if you have to hard-code your search results or add keywords to the page or whatever then you should do it so that, when people execute those searches, they’re guaranteed to get the information they were likely looking for. But you should also, then, look at those five search queries and say, “Well, why do people search for this? Should they have been able to find it by looking at some link on the homepage? Is it important enough that we actually need to have a link on the homepage that would take people to this information and that would also then show up hopefully in Google searches?”
“And you do it once a month. So next month you’ll probably have some of the same queries and some different ones.
Genius! I’ve been having trouble with some of the visitors Google sends, and this “compare to the SERPs” monthly review sounds like just the solution.
A third solution, which would probably be more precisely effective – since competitors are also only guessing at intent and may not be guessing well either, like a game of charades – would be to just survey visitors. “It seems this page is having trouble satisfying visitors. Would you mind helping us help you and others like you, by answering two short questions?
a) How did you get here? This can help us solve the problem for other people visiting based on a similar context.
b) What were you hoping to find or do here? The more detail, the better.
And that’s how you meet the needs of visitors that found a real page but didn’t see what they wanted on the page because the page didn’t have it there.