Visual search tools: The future of furniture retail?

Searching for products online is part of everyday life for most Internet users. The process often begins by typing in a search query. But change is afoot in online furniture retail: shopping using photos.

The online furniture trade is steadily growing. Whereas a trip to a furniture showroom used to be the norm, today it’s possible to order a new bed in just a few clicks without leaving the sofa. But what happens when customers want new furniture, but the conventional search function doesn’t deliver what they’re really looking for? It’s not uncommon for prospective buyers to turn first to Pinterest or Instagram – indeed, people spend hours searching for inspiration on these platforms. According to Pinterest, 60 per cent of its users have used the image-sharing social network to make decisions about home textile purchases. A quick hashtag search for “interior design” or “furnishings” returns billions of results in seconds. Photographs are fundamental to the process of reaching a purchasing decision. So how can furniture retailers ensure that customers find exactly what they’re looking for? This is where new software comes into play that will revolutionise the furniture trade, especially in the future: visual search. Because wherever verbal descriptions reach their limits, good images can provide clarity.

Although the terms “image search” and “visual search” may be similar, they actually take different approaches. With a visual search, the search query is no longer input using language – for example, by entering or saying certain keywords – but instead by uploading photographs taken with a smartphone or tablet. The visual search then delivers similar images. The technology uses artificial intelligence (AI) to break down the image into its component parts and thereby extract information about the main object. The pre-trained algorithm then compares this information with other objects that have already been scanned.

This new type of search query is of interest to many different sectors – from fashion and tourism to home furnishings. The mail order business OTTO introduced its own visual search app alike onto the market back in 2018, making it an early adopter in the furniture industry. Customers can photograph home furnishings directly from within the app or upload a picture to find similar or even identical items. The unusual thing about alike is that it doesn’t just search the OTTO catalogue but instead trawls through over 300 online shops offering more than one million products. The Swedish furniture giant IKEA was also quick to invest in visual search technology. In 2017, it launched IKEA Place, an app that even harnesses the power of augmented reality to allow users to project furniture into their own homes in real time . Features that make it even easier to furnish a home.

But what happens when customers don’t find what they’re looking for? Jens Krahe, Managing Director Creative & Technology at Plan.Net Cologne and an expert in the user experience, also feels there are still shortcomings in the technology: “The point of a search tool is always to provide the customer with results as quickly and easily as possible. The visual search function cannot yet do this on the same scale. For example, you’re sitting in a cafe and you like the green vase that’s on the table. But you’d prefer it in grey. If you start a visual search with a photo of the green vase, it will show you exactly the same thing: green vases. So, to get the results you really want, you have to add additional information manually – if that’s even possible. And from this point on, the effort involved in a visual search is relatively great for the user.” Thanks to Google, the true pioneer among search engines, customers are used to getting the right search results very quickly. That makes it all the more frustrating when the search takes an unusually long time or does not deliver the desired results. To get around this, furniture retailers and manufacturers have to prepare their product catalogues so that the algorithm works perfectly – an expensive project that is currently “more an additional feature and not yet applicable for everyday use”, says Jens Krahe.

Even if visual search technology is not yet a standard function that customers use intuitively, it still presents some great opportunities for furniture retail. The ability to search using images can be useful for customers, especially when looking for niche products or unique items. Products that are particularly difficult to describe could be found quickly using a well-trained visual search algorithm.

Jens Krahe believes this new way of searching could offer great opportunities for the furniture trade, especially in the future – but not in the form of expensive individual applications. Augmented reality is the key here. At its Worldwide Developers Conference in 2021, tech giant Apple revealed what harnessing this technology could look like in practical terms. With Object Capture, the respective item is photographed from as many different angles as possible. The software uses these photos to create a detailed 3D model, complete with texture. “That would be one possible way of making visual search work better. Then it really would be a visual search,” says Jens Krahe.

However, delivering this process requires a huge amount of data, which needs to be formatted perfectly, so it’s not yet commercially viable for many furniture retailers. “I would wait until visual search in the form of augmented reality becomes an operating system feature of its own or is built in to major search engines like Google. Because many more devices will then have this feature and it will run like a normal Google search, only with images.” This would significantly reduce the effort on the part of customers, because they would no longer have to download the apps of individual furniture retailers and manufacturers and search each of them using images. Instead, it’s conceivable that they could access the catalogues of all online furniture retailers via a central platform. It’s only then that the true potential of visual search technology would be revealed – once running a text-based search requires greater effort than a visual search, the application will be much more attractive to customers.

Visual search is definitely a trend that furniture retailers and manufacturers should monitor. Companies that ensure their product catalogues are properly formatted at an early stage stand a chance of coming top in the search results and being at the forefront of the furniture industry when the time is right. This is also the conviction of technology expert Anna Lukasson-Herzig. With her start-up Nyris, the Berlin native has specialised in visual search. In an interview with ambista , the founder reveals what the future of search queries looks like and why visual search will be an indispensable part of it.