Why your internal site search is probably lying to you about what customers want
If you've ever looked at your site search analytics, you've probably seen a tidy list of the most popular search terms. "Blue widgets" gets 500 searches a month. "Red widgets" gets 300. Excellent, you think. People clearly want blue widgets. Job done.
Except that's not really what the data is telling you at all. What you're actually seeing is a list of things people typed into a box. You've got no idea whether they found what they were looking for afterwards. For all you know, everyone who searched for "blue widgets" immediately gave up and went to buy them from your competitor instead.
This is the fundamental problem with most site search analytics: they measure activity, not success. They tell you what's happening, but not whether it's working. And that distinction matters quite a lot if you're trying to work out what your customers actually want.
The difference between popular and failed searches
Here's a thought experiment. Imagine you run a bookshop website, and your analytics show that "science fiction" is your most popular search term by miles. Brilliant news, right? Your sci-fi section must be flying off the digital shelves.
Or perhaps everyone's searching for science fiction because they can't actually find it any other way. Maybe your navigation is a mess. Maybe you've categorised everything under "speculative literature" because someone in marketing got a bit carried away. Maybe your sci-fi section is buried three clicks deep under "Books > Fiction > Genre > Speculative > Science Fiction" and nobody can be bothered.
In the first scenario, you're doing well. In the second, you've got a serious problem. But in both cases, the search volume looks exactly the same.
Failed searches are queries that don't lead anywhere useful. They're the digital equivalent of asking a shop assistant where something is, being pointed vaguely in a direction, and then wandering around for ten minutes before leaving empty-handed.
How to spot a failed search
The good news is that failed searches tend to have fairly obvious characteristics once you know what to look for. The bad news is that most analytics platforms won't highlight them for you automatically, so you'll need to do a bit of detective work.
Look for searches where people don't click on any results. This is the most obvious red flag. If someone types something in, looks at the results page, and immediately searches for something else or leaves entirely, they probably didn't find what they wanted.
Check the bounce rate for search result pages. If people are landing on search results and then vanishing, that's a problem. A certain amount of bouncing is normal—sometimes people are just checking if you stock something—but if your search results have a much higher bounce rate than the rest of your site, that's worth investigating.
Watch for repeated similar searches from the same user. When someone searches for "laptop charger", then "laptop power cable", then "laptop power supply" all in the same session, they're not suddenly overcome with curiosity about synonyms. They're not finding what they need and they're trying different words in the hope that one will work.
Look at searches that return zero results. This one seems obvious, but you'd be surprised how often it gets ignored. If people are regularly searching for things you don't have, or things you do have but call by a different name, that's extremely useful information.
The zero results problem
Zero results pages are particularly interesting because they represent the purest form of failure. Someone wanted something specific enough to type it into your search box, and you've told them definitively that you don't have it. Even if you actually do.
Sometimes this happens because you genuinely don't stock what they're looking for. Fair enough. But more often, it's a terminology problem. You sell "trainers" but they searched for "sneakers". You've got it filed under "RJ45 ethernet cable" but they typed "internet wire". You offer the service they want, but you've described it using internal company jargon that no normal human would ever think to search for.
The solution here isn't necessarily to change all your product names—though sometimes that's exactly what you should do—but to make sure your search understands synonyms, common misspellings, and the various weird ways people might describe what they're looking for.
This is also where your search analytics become genuinely useful for product development. If a hundred people a month are searching for something you don't actually sell, and zero results is the best you can offer them, that's a pretty clear signal about an unmet need.
What good search analytics should tell you
Rather than just showing you search volume, your analytics should be telling you about outcomes. What happened after the search? Did people click on results? Did they click on multiple results, suggesting the first one wasn't quite right? Did they end up making a purchase, or booking a call, or downloading something, or whatever your site exists to make happen?
You want to be able to see your searches ranked not by popularity, but by failure rate. Show me the searches where people looked at the results and immediately left. Show me the queries that led to clicking on three different products without adding any to the basket. Show me the terms that people search for repeatedly because they're not finding what they need.
Most analytics platforms can give you this information, but you'll need to set up some custom tracking or spend time filtering the data. It's worth the effort. Once you can see which searches are actually working and which ones are failing, you've got something much more useful than a simple list of popular terms.
What to do about it
Once you've identified your failed searches, you've got a few options depending on what's causing the problem.
If it's a terminology issue, add synonyms to your search configuration. Make sure "trainers", "sneakers", "running shoes", and "kicks" all lead to the same place. Include common misspellings if your search doesn't already handle them automatically.
If people are searching for things you do have but that aren't showing up in results, there's something wrong with your search indexing. Either the content isn't being indexed properly, or your search algorithm is rubbish at matching queries to relevant results. This usually requires either better tagging of your content or a more sophisticated search tool.
If searches are returning results but people aren't clicking on them, look at how you're displaying those results. Are the titles and descriptions clear enough? Are you showing the most relevant results first, or just the newest ones, or the ones you're trying to shift because they're not selling?
And if people are searching for things you genuinely don't offer, you've got a decision to make. Is there enough demand to justify adding it? Can you point people to something similar? Or do you need to find a way to set expectations earlier, so they don't get as far as searching for something you can't help with?
The underlying point
Site search is one of the few places where people explicitly tell you what they want. They're literally typing it in. But that information is only useful if you pay attention to what happens next. Did they find it? Did it help? Or did they wander off disappointed, adding another data point to your list of popular search terms that looks impressive but means nothing?
Most sites treat their search analytics as a curiosity—something to glance at occasionally when putting together quarterly reports. But if you dig a bit deeper and focus on failures rather than just volume, you'll learn far more about what your customers actually need and whether you're providing it.
Your search data isn't lying to you, exactly. It's just not telling you the whole story. The interesting bit isn't what people searched for. It's what they didn't find.
We Know It's A Lot to Take on Board
Learning the ins and outs of Console and Analytics can be a brain ache at times (and yes, that's an understatement). Which is why if we build your website and you pick our ongoing management service, you don't have to worry about it because we look at all the data for you. Not only do we write the content for you, we also keep an eye on which articles are getting impressions, which could generate more clicks and act on it for you.
To see how we build your website and handle all your content then take a look below:
But if you would like to learn more about Google Console and Analytics, we have lots more insights you can read:
The Canonical Doppelgänger Dilemma: How to Tell Google Which Page is the Original
Clicks vs. Impressions in Google Search Console: Reading the Performance Tab Without Drifting Off
What does it mean when people search for something on my site but don't click any results?
It usually means they didn't find what they were looking for. Either your search results weren't relevant, the way you're displaying them isn't clear enough, or you're not showing them what they actually wanted. This is called a failed search and it's a sign something needs fixing.
What should I do if people keep searching for products I don't actually sell?
First, check if there's enough demand to justify adding those products. If lots of people are searching for something consistently, that's a pretty clear signal there's a gap worth filling. If you can't or won't stock it, at least you'll know why those searches are failing.
Why do people keep searching for the same thing multiple times on my site?
They're not finding what they need on the first try, so they're rephrasing their search hoping different words will work better. If you see patterns like 'laptop charger' then 'laptop power cable' then 'laptop power supply' from the same person, your search isn't understanding what they want.