From Lego to iPhones, here’s what our customers search for most

Customers browsing our shops type a good 500,000 unique queries into our search bar every day. But what do these 190 million annual searches reveal about the people typing them? Which products are our customers most hyped up about? When do most searches come flooding in? Which queries come up empty? With several teams of developers working on our search feature every day, we’ve asked them for some exclusive insights into our customers’ search patterns.

Order your new ski goggles over après-ski drinks

«Our customers don’t want to root around for the product they’re looking for. They want to find what they’re after quickly and easily,» says Ryan Prater, Product Owner with Digitec Galaxus’s Team Endeavour. The search process has to be speedy – especially at the weekends, when people are out and about. If somebody searches «ski goggles» while they’re sitting in the après-ski bar, for example, they’re keen to wrap up their order ASAP so they can get back to drinking their «Schümli Pflümli». «Our data shows that our users enter five per cent fewer characters into the search bar via mobile devices on Saturdays than via desktops on Wednesdays,» Ryan adds. With more and more people using our mobile app, this trend’s set to become even more pronounced.

Clara Goebel, a Data Analyst on Digitec Galaxus’s Team Nightingale, who works closely with Ryan, explains: «The complexity of the data allows us to understand in great detail how users interact with the site. This means we can optimise our search feature and user experience in a targeted way.»

Making the search feature efficient is a highly complex job, requiring numerous developer teams all working together. Team Endeavour’s dedicated to user experience, focusing on how users interact with the search bar and filters. Meanwhile, Team Nightingale supports them with extensive data analyses. Other teams, such as Nautilus, take care of search algorithms and rankings or make sure the site navigation’s running smoothly.

Apple and Lego dominate the search rankings

Some long-standing favourites crop up in the search stats, most notably Apple and Lego products. These are very popular with our customers, with Lego only briefly giving up its position at the top of the rankings in recent months. On last year’s hot days, searches for the word «fan» stole the top spot. We also saw frequent searches for power banks, backpacks and vacuum cleaners.

As well as these evergreen favourites, it’s also worth looking at search trends. These reveal which products and product categories are catching our customers’ interest at a given moment, and are always shown as a comparison to the previous week. Searches like these go up and down depending on the season, upcoming public holidays or our latest magazine articles. When the weather began warming up in June, customers searched «dyson fan» 275 per cent more frequently on average. Events such as Halloween are also clearly reflected in the search data. Searches for «halloween costume» went up 101 per cent in October. When searches for a particular term shoot up compared to the previous week, it’s usually because a new gadget has been released or our editorial team has published a review. That’s why we saw a rise in searches for «9800X3D» around the gaming processor’s release in November, and an influx of new searches for «sex toy» when our annual Sex Report came out.

One in ten searches come to nothing

Not all searches bring about the desired result. In fact, eight per cent of all queries failed to generate a hit in 2024. Thanks to a new search algorithm launched in November, however, we’ve managed to bring this figure down to two per cent. Previously, when a customer searched for a product not stocked by Galaxus, the search would generate zero results. Since the new algorithm was introduced, these searches have started bringing up the most suitable alternatives.

Although «deal of the day» is a particularly common search term on our website, the search used to come up with nothing. As of mid-December, however, customers searching for «deal of the day» have been presented with the day’s bargains. People often search for AirUp water bottles too, but Galaxus doesn’t carry the German brand’s bottles in its range. Our search teams keep our buyers in the loop about which searches fail to generate results, so that the products in question can be added to our range. This is exactly what happened with Scrub Daddy sponges, which we now stock.

Other factors influence zero-hit pages too. In Switzerland, for instance, only seven per cent of search queries produce no hits. In Germany, however, the figure’s closer to 12 per cent. The reason? Availability differences between countries.

Data-based decisions: how our teams test new concepts

«Given we have this much data, and that we want to focus on improving customer experience, it makes more sense to try out ideas rather than investing months of time in creating a road map,» Ryan says. Team Endeavour’s currently running an A/B test for automatic filter sorting on product pages. Ryan explains: «If you’re looking for a bike chain, your first question is going to be, ‘How many gears does the cassette on the back wheel have?’ With this in mind, the cassette compatibility filter should come first.» The team’s hypothesis was confirmed, and thanks to the experience gained from the A/B test, the feature was rolled out to all our shops’ users.

Tests are also being carried out to establish whether AI can be incorporated into our search feature. Team Nautilus is currently working on an A/B test that’ll allow us to understand longer, more complex search queries better. For example, if someone searches for «metal component for cleaning a drain», the AI could interpret what the search means and guide the customer to the right product. Prater says: «We usually run these tests for two to three weeks to achieve statistical significance.» This enables the teams to quickly get meaningful results. The users are divided into two groups at random; one tests the new feature, while the other serves as the control group and sees the usual version. The data collected is completely anonymised and allows the developer teams to draw conclusions about which changes improve customers’ search and shopping experience.

Our teams will continue investing a significant amount of time in helping our customers find the right products in future. What do you think of our search feature? Where do you see room for improvement? We’re looking forward to reading your opinions in the comments.