For 2 decades Google has provided a search engine that is accessible to all. Every year, the company raises the bar with search innovations like voice search and search using your mobile device's camera.
Google also recently announced that they are testing the integration of artificial intelligence into search engine results, which will open the door for a more natural way of searching. As this will allow the user to directly ask their questions and get an answer that is compiled from content on the web that Google thinks is credible under the O-E-A-N (E-E-A-T) system.
In this GrowEasy article, we'll take a look at the major search innovations that Google has integrated or will integrate this year as a major update. It, which will contribute to the algorithm update in March and will prevent the annoying to prevent "spam "a in search engine results.
Fencing (or highlighting, or strikethrough) for search
Google's idea with this new search method is that it makes it easier for users who want to search for something they've seen in an app outside of the search engine (e.g. YouTube, Instagram, Facebook, etc.) without separating them from the app itself. Which consequently eliminates the need for the user to have to interrupt their experience in the app they are in to search in the search engine.
Here's a very useful application of the feature:
You see a product and show a lot of interest. Here, instead of wondering how to describe it with keywords or the "traditional" way of searching for the specific online store, you put your finger around it and
Useful applications of this feature are the times when you see a product you are interested in and instead of having to go into a search engine and wonder what to type to come up with the online store that sells what you saw, you directly fence and Google finds what you are looking for.
Google will also continue to recommend Google Ads in these types of searches coming from fencing.
Unfortunately for now, this feature is only available on the phones: the Pixel 8, Pixel 8 Pro, and the new Samsung Galaxy S24.
You can see a demo of this feature in a video provided by Google here.
Point your camera and get a response
Something that's not new to Google is getting search results with a photo.
To take advantage of this unconventional feature, which although has been around for a long time, is still gaining popularity you need to open the Google Chrome app on your Android. Then select a new window and you should see a camera icon in the search engine.
If you click on this camera instead of starting to type in the search engine, a completely new window will open. Right now you see Google Lens, where you can choose between using your camera or an already-photographed file to search the search engine.
Additionally, you can scan the text from your photo and search for it directly as a simple Google search with the "Homework" button or translate it with Google Translate with the "Translate" button
Combined search
A feature of Google search that most people don't know exists at all is the combined search. K, which is when you use the image search and add a keyword to influence the results that come up from that search.
For example, you saw a pair of shoes you liked, used them, and decided to Google them to find out their price. You open Google Lens and search for the shoes with a photo you just took. You look through the options and realize that these shoes, if they were only in black, would be something you would gladly purchase. All you have to do is type the keyword "black" into the search engine next to the picture and you will now get results of the same shoe, only in black.
The disadvantages of combined search and photo search
Searching with photos and additional keywords is very useful at first glance, until we tried it in practice.
What impressed us was how Google's algorithm recognized the completely random items we selected and found relevant results for most searches.
But what made it difficult was when the photos were taken in "bad" conditions. For example, when the light wasn't sufficient for a good quality photo, the colour of the object would blur and then Google would recommend completely different objects, which nevertheless matched in shape with the one we had photographed. Which shows that the algorithm is doing its best to produce relevant results even with "not so easy to interpret data".
Another flaw we noticed completely by accident was that objects without much prominent color and relief worked very well with Google Lens. H, but when we added a keyword to some of them, the results were quite confused.
For example, when we photographed an eyeglass case that stood out with almost no features that would make the algorithm easier, Google was able to find the exact same case, but when we tried searching for the same photo with the keyword "red" expecting to see a red case, Google recommended something completely different. What we saw on the first page was a red ICOS, which incidentally was extremely reminiscent of the shape of the case.
Combined search is a great technology that should NOT be taken at face value though. Our advice is to take advantage of the functionality because it can still make your daily tasks much easier. But still, don't completely trust these searches completely.
What to expect in a search engine
Google is introducing its latest technology that allows the user to ask direct questions related to the image they are searching for in the search engine and be able to get an answer from the AI in their search.
Google offered the following situation as an example. You are browsing items at a garage sale and see a role-playing game that has been left on the table with no box or rules. You pull out Google Lens, take a picture, and it immediately tells you the rules of the game, how to win, and where you can buy it.
Google's ambitions to integrate artificial intelligence into search results could completely change the way we search for information on the internet. But will they tackle challenges such as the high cost of supporting the many queries to artificial intelligence or insufficient data for quality training? We can only guess, but for now, such integration is too big a risk for the company.