Google Lens: SEO for images on another level

Not all the new features that Google launches have the same impact. Some simply stay there without contributing much, but suddenly...they can become a way to change everything.

This is the case of Google Lens, an AI-based image search system that has been working for years in the mobile search engine app, but more as a curiosity than as a way to find a real alternative.

The truth is that there are movements that make us think that the bet for Google Lens is going to be strong in the next months. And it is worth seeing what we can do to optimize SEO based on this.

One of the "classic problems" of search engines is that, while they increasingly understand written language, they do not have an efficient way of interpreting what appears in a photograph. For SEO purposes, we have always had the resource of alt text and little else.

Being aware of this weakness, and relying on the advance of technology, Google decided to create a system that, based on Artificial Intelligence, would be able to know what an image represents.

It is a way to convert those pixels that are something abstract into something that can be understood and that is the same as what the user sees.

This feature, which was previously inside the Google search assistant app, has been available natively on the phone's browser for some time by clicking on a square icon in the upper right area.

If we upload or take a picture of a bag or a T-shirt, Google will know that what we upload is a bag or a T-shirt and that we want relevant information.

The difference with Google Images' inverse search is that, in this tool, if we upload a photo of a brown bag, it will understand a color pattern and a composition. By this, I mean that it can show photos of handbags, but also of a sofa or a dog.

It does not analyze the content of the image, as Lens does.

The evolution of SEO towards alternative searches is clear. We have been seeing how search by voice has been growing steadily for years. 

This is, to a great extent, thanks to the weight that the cell phone is getting in all our day to day. And, although to a lesser extent, by IOT-connected devices, such as smart speakers with built-in voice assistants, televisions, smartwatches…

Another of the native features of phones is, precisely, the fact they have camera. This means that the user may, at any time, want to find something on the Internet that he is seeing with his own eyes or searching directly on Google Images

Until now, this situation meant an additional effort for the user, who had to be able to convert that visual reference into a search intent. Therefore, they had to convert it into keywords -and there, a lot depends on factors such as the user's ability to do so, their education, level of knowledge of what they see, cultural context...

With Google Lens, things can change radically and be noticed especially in eCommerce.

Imagine that a person sees a friend who is wearing a T-shirt that he likes. It would be very difficult to find it looking for: "black t-shirt with an orange stripe on the left sleeve". Not that it is longtail: it is that nobody in their right mind would optimize a page for this.

Google Lens, on the other hand, scans the image and compares it with its hugeindex of photographs searching for exactly that shirt and different alternatives that are similar to it. 

All these Artificial Intelligence tools are more complex in all senses, including that of organic positioning.

Anyway, there are 4 basic ones that we must have covered if we want to really have our eCommerce ready to rank by searches in Lens.

Logically, the number of images we have available will expand the possibilities of matching with what users are looking for.

It gains special weight to make our own good quality photos and not use only those that the manufacturer distributes. In this way, we differentiate ourselves from the crowd and create our own search assets.

I am not talking about an overprinted watermark. That is uncomfortable for the user and creates a bad user experience. Try to make it visible on the product images because it is going to help you make a difference.

If you have a physical or retail store, do not forget to upload images of it to your web where your logo is also visible. This is particularly important for local SEO.

This is one of the basics of SEO that does not go out of style even in times of SEO for Google Lens.

It is clear that Artificial Intelligence is making great strides, but Google always continues to appreciate these signs to understand the content of photos and make them clear when in doubt.

In addition, alt text is important for accessibility. Certain visually impaired users need such support when navigating. 

Therefore, it is a good practice to rename each photo (keyword.jpg instead of 3456544_DSC.jpg) and always use the alternative texts. 

All photographs include a series of meta-information within the archive itself. It is a data file that refers to the model of camera, time the picture was taken, place...

This data is sometimes removed from images because it does not add anything visually and increases the weight of the file. But... And for SEO? Some time ago, the controversial Matt Cutts made some statements saying that "they could be a positioning factor".

Years later, with Matt out of Google's public eye, we see that this was the idea.

The good thing is that EXIF data can be edited with a minimum of knowledge and a program like Photoshop or GIMP. This way, we can include our keywords.

Will Lens be a revolution? It is difficult to know to what extent, but it is sure that it will go further and it is worth taking into account and adapting.