Google held their latest Google I/O event on the 19th May 2017, an event where techs and the search industry wait with baited breath to see what innovation is coming from the Mountain View giant. Amongst the many topics covered in the keynote there was one which got us both intrigued and excited – Google Lens.
Google lens essentially adds some extra smarts to the AI and computer vision technology that already exists. It packages it up to create a mobile friendly image search application that applies context and understanding to what your camera sees.
The search giant has been able to perform reverse image lookup for quite some time (but not easily from a mobile device), it has also been able to interpret images as well, think about the fact you can use google translate to translate foreign language signs etc. into your own language, that’s pretty clever!
However what Google are proposing with lens is that you use the image itself to search for something or add something to a calendar, set a reminder with google assistant etc. That image has context associated with it to be able to return relevant rich information for the user. Examples they offered during their keynote included taking a picture of a plant or flower and Google returning a range of information from what the flower is called, how to care for the plant etc.
Or how about taking a snapshot of a restaurant before you enter and instantly getting review information?
This is all really useful information that ties directly into the moment and micro-marketing mindset, but what does it actually mean for us in the search industry and how can we prepare?
Google have, over time, started to move towards a single answer search function. If we look at the rise in home assistants such as Google Home, Amazon Echo, the announcement of Apple’s Siri equivalent many of these devices and functions don’t have the ability to return a page of results to the user, they return a single result. Google started utilising featured snippets that try to return a single answer highly relevant to a searches query at the top of results and whilst this new function has potential to return several results or resulting actions it still drastically reduces output from more traditional ‘search’.
With that in mind there are a couple of elements in a search marketer’s handbook that become infinitely more important:
- Image optimisation
- Structured data
How do we optimise our images for Google Lens?
To enable Google to return relevant result from an image reverse lookup perspective the associated images will be key. Therefore ensuring your images are fully optimised could be a very important factor so think about all the elements that you can impact on your own or your clients’ sites:
- Filename
- Alt attributes
- Image titles
- Image captions
- Image sitemap.xml files
- Local business image submissions
- Inclusion of product images in review content
- Application of schema for images (ImageObject)
- Application of schema for content supporting the image or vice versa (e.g. LocalBusiness)
Should you or your clients be interested in Google Lens?
Absolutely! Think about use cases where you might have a need or desire to appear off the back of an image based search:
- FMCG brands – someone takes a snapshot of your brands bottle do you have recipes schema in place for cocktail recipes that include your drink, or a food based recipe, maybe they are looking for more detail on nutritional aspects.
- Clothing & other retailers – someone takes a snapshot of an item of clothing they have seen on someone, in a magazine or something in a shop window they can instantly get reviews and suggestions where to buy as well as more detail about the garment
- Local businesses – someone takes a snapshot of a restaurant and wants to make a reservation for later, or see your menu without physically going into the premises
These are just a few examples of where Google Lens might come into play and the extent to which this technology influences search is yet to be seen, but it’s always beneficial to be prepared and stay ahead of the search curve!