Image Search for eCommerce websites is becoming increasingly important

Roxana says: "Images are now playing a bigger role in product search. If you go on Google and search for one of your best-selling non-branded product terms, you might see some of the bigger brands competing against you. If you go and do the same search on Image Search, you might be surprised to see that some of the names appearing are not the big brands you were expecting. Some of the big brands are not optimising their images as well as they could."

Is it important to have unique images on your website? I would imagine that a lot of eCommerce brands are taking images from feeds, so there could be hundreds of other websites using similar images.

"The search engines are ranking the page, not the image itself - even though it's the image showing up. You could use the same image as somebody else, if your page is better optimised for that image. It doesn't matter if your image is not unique. Just make sure you're optimising for that image better than somebody else.

The more websites that use the same image, the harder it gets to rank for it because competition is bigger, but it also depends on how the image is used. If everybody's using the exact same product descriptions and everything then, in this case, it will be hard to outrank an original website with the same image, and the same text, regarding that product. However, if you're using an image in one context, and somebody else might be using the same image in a different context, then there isn't any competition because you're going to rank for different things in the first place."

Is Image Search traffic increasing, and by how much?

"Both Google's and Bing (which has an extraordinary image search platform as well), have been investing a lot in bringing all these different features to Image Search. It's hard to say exactly what percentage of total search it accounts for because there isn't as much data around image search as I would hope for. There's an older survey, done by Google in 2019, which showed that 50% of eCommerce users used Image Search to help them choose what they want to buy. That was three years ago - imagine what it is now, with more people using mobile phones. It's so much easier to look at images than actually read text on a page on your phone."

It's obviously part of the buyer journey. Is it used at a research phase, when people know what product they're looking for but they're not necessarily sure what it's called?

"It is. People have so many different intents and reasons why they would use Image Search. They might just be looking for ideas, so they can upload a picture of a specific pattern to find a bag that uses that pattern. They might be looking for a specific style. If you're looking in a magazine, you like a combination of a pair of jeans with a blouse and you think, 'I have jeans like that, now I just want to find a similar blouse I can buy.' you're going to do more of an image search than a text search.

There is also the situation where you know exactly how the product looks, and you just want to find who's selling it, and maybe for what price. Then there are people who are just ready to make a purchase. They know exactly what they want - they just want to find the place to buy it from. There are many different scenarios where you would use Image Search."

Is Pinterest part of this image searching matrix, or is that an entirely different proposition?

"Pinterest shows up a lot in Image Search. I don't know if this is a good thing, because eCommerce websites are now competing with Pinterest. I think search engines, especially Google, are smart enough to know about the intent of the user. They can tell whether somebody is just browsing images for ideas, so they're still researching, or when people are ready to buy. The intent in their queries changes - and this is where Pinterest doesn't show up as much."

What are your general tips on file type, size and positioning, additional text, and where that text goes, when optimising for image search?

"Firstly, if you want your image to rank well, you have to give the image the importance it deserves. That means you should place it above the fold, or as high on the page as possible. If the image is not the main focus of the page, chances are you won't rank well for that particular image. It's all common sense - that's all there is to it. It's not one of those SEO techniques like putting your keyword first.

Next, your images have to be large. For example, Google recommends people use images that are 1200 pixels or more if they want to do well in Google Discover. The same recommendation works on Image Search - the larger your image, the greater the chance you have to do well.

I know large images can take up hundreds of kilobytes, which is a problem for PageSpeed, and everybody will feel the urge to compress the images into oblivion. Please don't, because sharp images tend to do better in Image Search. I always go with the lossless compression, although unfortunately, this is not always satisfying, and the image file can still come out too big.

If you're worried about PageSpeed, I'd suggest considering using responsive images instead, which is just a different way of adding images in HTML using the picture element. It allows you to give users different image sizes based on their device, and you can even have different degrees of compression. For example, you can have a smaller image for mobile. There, you don't need it to be very sharp because it will be very small and you're not going to see a lot of detail - so you can compress more, but you still give search engines access to a large and high-quality image file.

This picture element has about 96% global adoption by browsers, which means it should work for most traffic. Internet Explorer and Opera don't tend to do well with this, but every other browser works perfectly fine.

You can also look into using different image formats. For instance, you can use WebP, or a newer format called AVIF, and both can reduce the file size even further - without affecting the quality of the image. Just check browser compatibility. WebP has a 95% global adoption, but AVIF is still quite new and only covers about 70% of the browsers and browser versions that are currently in use."

What are the benefits of using WebP and AVIF instead of JPEG?

"The main benefit of WebP is you can get about a 25% compression just from the file format, without losing the quality you would normally lose if you did normal lossless compression. I haven't tried AVIF yet, but you're supposed to be able to compress it even further without losing the quality. Once more people use it, we might get better numbers on how good it is."

What about dimensions - ideally, would you be aiming for a square image?

"That is a very good question because I haven't tested it, and I don't like to talk about things I haven't tried and only heard about. Apparently, Google will prefer certain dimensions of images, such as square, or rectangular in the 16:9 ratio. This feels logical because it's easier for Google to show as much of the image as possible in the pre-set image sizes shown on image search results. If an image is really important, would Google not show it just because it doesn't have the right ratio? I'll let you know after testing."

Joy Hawkins shared a tip to include a square photo near the top of a page. So that's for the conventional SERP because Google is starting to use images as part of mobile search results next to the actual blue text listing.

"In web search, in the conventional SERPs, you can even have smaller images because they're mostly used for mobile anyway. Google doesn't care to have these big, quality images. If they are square, it's easier for Google to show them because you will know all the information in the image will be visible, and they don't need to cut it off. It definitely makes sense to use square images - if that's what you're targeting."

What about the quality of traffic from Image Search? Does it tend to convert just as well or is it a different stage in the buyer journey?

"It depends what you use Image Search for. In stock photography, for instance, we convert way better in Image Search than we do in Web Search. That's normal, because people who want to buy a specific photo will use Image Search to find it. They also use Web Search just to find websites that might have that specific photo.

For eCommerce in a classical way, I assume it's going to be the other way around. When people go to Web Search, they want to find the websites where they can buy things. That's going to have a good conversion. When they go to Image Search, they might just be in the research phase, trying to find a product in the way they were imagining it and seeing which website sells it. They might end up actually clicking from Image Search to that website and end up converting. It should still be a pretty good conversion rate from Image Search, and if it isn't - then something's not working. Maybe you've got the wrong images in there or you're showing up with your blog and not actually with your product."

If an SEO is struggling for time, what could they stop doing to focus more on Image Search?

"There are many SEO things we spend time on that don't necessarily have that much return on investment. One of the most recent examples is that we spent so much time as an industry optimising our titles for our pages. Then Google just went and rewrote those however they wanted - proving there's no point wasting a lot of time doing that. I'm not saying never do it again. Maybe just take a break from this activity and optimise some alt texts, or some image file names to make them more descriptive and help those images do better."

You can find Roxana Stingu over at Alamy.com.