Re: What I have learned by posting 10600+ photos, earning 8300+ points and getting 130+ million view
Hi
I think that there is no algorithm based on the content of the picture and no decisions on photo merit can be made by anyone at Google because they would have to employ a population to do it.
It's I'm sure a crowd-sourced algorithm based on the things people want to see, based simply on them looking over I guess something like the last couple of days - in your example the first one is good but the second shows far more of the blackboard menu - that's part of what people want. That would not work anywhere else than a place to eat so no advanced photo-analysis could pick what's needed for a certain type of location.
@SteveBradburyI would love to know for sure. I have pictures that have made the top listing and just stay there. My opinion is the algorithm picks the photo and until a better picture is posted, based on the algorithm shows up for that listing it will stay there. Some places just have huge amounts of traffic. Resturant cash register photos get views. People search for places a lot. I often doubt the numbers are real because for some locations they are just nuts high. The photo linked to above as of today has 361,000 views since last year and was the cover photo for the listing. Now the current one is similar but different. I do not go back to monitor the photos enough to track what is being displayed and know the views vs click but certainly, Google does. I would say they do test it. In this example how would you know why they bumped mine? I do know this if you take pictures like this you will get views. If you take pictures like this in places that get a lot of traffic or in high population areas you will get views. I think the biggest view factor is if you get the cover photo you get views. If you get cover photos in high population or transient visitors you get more views. The key is to watch if your photo is the cover shot since that skews the number so much. Food gets views, we all eat. In this digital age, the hunter-gatherers are Googling it. This picture was the cover photo got view fast then identical store down the street 2 blocks the same photo set up got very low views. My point is It is hard to see the whole picture but these types of photos get high view count.
Re: What I have learned by posting 10600+ photos, earning 8300+ points and getting 130+ million view
Hi, Steve @SteveBradbury I respectfully disagree "I think that there is no algorithm based on the content of the picture and no decisions on photo merit can be made by anyone at Google because they would have to employ a population to do it". There are videos describing how Google can read and use Streetview to find company names and map that to the location. A look at this photo API description shows some of the capability. I believe that it is all automated. The ranking may change over time based on user interaction but also based on what is added by other users so it is hard to know exactly what is happening unless you watch and document each picture on the particular listing.
Re: What I have learned by posting 10600+ photos, earning 8300+ points and getting 130+ million view
I have read so much of this post and seeing all the AI capability, the concept of "pic wars" and the strong desire for headline photos by so many people make me happy and realise I don't do these habits alone. I've reached a quarter of a billion views and I'm eager to share my entire workflow with the LG community.
I might format it in a similar way Svein has, but I'm really willing to share my journey of how I got to where I am now. Soon enough, I'll receive the congratulatory Google email (most likely by Wednesday) and from there I'll form the basis of my post.
The rankings of the photo change pretty regularly, I've usually seen my photo getting knocked off the top spot before I find out that another photo of mine has taken its place months later. So the AI is constantly changing and evolving to see what photos catch the eye of data consumers, what influences their decisions to click through links or go to the store, and so on.
I know for many clothing stores, train stations, restaurants, Subways and Pizza Huts (yes, these are quite a gold mine for views) I've found almost the magic recipe to hit the top spot each time I upload them, but these are secrets I want to share to everyone. I know a local guide has been closely looking at my contributions for a while (when I uploaded a photo that got 60,000 views a day, the LG decided to upload 30 photos of that listing to try and snipe the top spot, and did successfully), but utility and consumer data should be harnessed to the best of their ability, and we as Local Guides can be the catalyst of such.
I'll be more than happy to share it, one year six months after Svein's original post.
2022
