Hope springs eternal? To be frank I don’t see the synergy between the two yet but given Flickr was in a death spiral this should mean a bit of a reprieve and a chance to make another go of it.
Strange things afoot at Flickr. They seem over run with porn spam. Just as staff solved the issue of spammers using mentions to draw people to porn ads, there are now bots creating dozens if not hundreds of new accounts a day to fill groups with more pornography ads. Then there is another spate of sites using the Flickr API to scrape people’s images and display them in unauthorized ways. Stats are borked and so are notifications for many folks.
It is not all doom and gloom though. The Flickr Blog top 25 seemed to feature some amazing images this year and they got there by more normal merit than hidden Flickr curation… although still some of that detected.
Well, I hope 2018 is a better year for Flickr friends than 2017. Either way… Happy New Year!
Last year I vigorously complained that Flickr was routinely featuring photographers who were blatant Contact Collectors. This was made patently clear in the 2015 selections for “Flickr’s Top-25 Photos in 2015” where many of the featured photos were from accounts with more the 50,000 follows and followed.
This years annual feature is called “Top Flickr Photos from 2016“. While Flickr has finally taken action against many contact collectors an odd pattern to those featured was noted again this year. This is what Flickr says they used to determine the selections for 2015:
To compile this list, we started with an algorithm that calculates a combination of social and engagement metrics, e.g. how often the photo had been faved and viewed. To avoid the results being merely a popularity contest, the selection also involved curation by Flickr staff.
This is what they said for this year selections (2016):
We’ve compiled this list based on a number of engagement and community factors. The photos were scored by an algorithm that calculates a combination of social and interactive elements, including how often the photo had been faved and viewed, among numerous others.
To most those two statements say essentially the same thing. So how did the stats of these two groups end up being so different? In 2015 most of the top images only have on average 200,000 views. Yet in 2016 most of the featured images have about 2,000,000 views. That’s a ten fold difference!
The difference this year is Flickr’s implementation of the new Feed. In the feed many users are confronted with non-contact images provided by Flickr (Featured and Recommended). This is what Staff have said about these images that get pushed to users.
There are two types of Recommendations at present. Those based on your interests. This is dynamic and based on your use of the site.
Then there are several hundred Featured Photographers that Flickr staff has selected based on our subjective view of interesting photography. We show a selection of photos from these photographers in various places on the site. Historically this has been to new account holders who sign up and haven’t yet followed anyone. We recommend that new members follow the Featured Photographers because we think those folks would be interesting. We also keep the list current, so if people take a Flickr break, they will be removed. We want to be sure new members of the community still have fresh content.
In the new feed, the recommendations are a blend of the two types, with the first being the most frequent type of recommendation if you’re not new to the site.
Research has shown that this year’s selections for Top Photos were drawn from accounts that Flickr staff had selected for featuring across Flickr and in particular the winning images had been featured or recommended in the feeds*! A single day as a recommended photo offers on average 200,000** views!! Without being selected by staff for being included as a featured or recommended photo an image had no chance of being selected by the algorithm since, as admitted by staff, views are a major factor in selection.
- * verified by research
- ** verified by queries
So not only do users have to worry about other users gaming Explore and Search but also Staff gaming other aspects of the site. 🙂
As this Help Forum post highlights once again… Flickr is not reviewing new accounts for their Safety Level. In the past not having your safety level set meant you would be excluded from search results, Explore and have limited visibility in Group pools. However, at about the time Flickr seems to have stopped reviewing and setting the safety level of new accounts it was discovered that new accounts were getting in Explore and were being featured in Tag search results. Unfortunately they still get excluded from all other search results and still have limited visibility in group pools.
Instead of acknowledging or addressing the issue, Flickr has let this drag on for months despite numerous queries in the HF. Why would they not address this? The only reasonable explanation is staff levels have been trimmed so much that those few remaining are just a skeleton crew trying to keep things together until the sale to Verizon is complete. This lack of staff resources would explain why many other major issues are also not being addressed.
The question then becomes is Flickr clinically dead and only on life support until someone decides to pull the plug? What’s left for users to do? Pray perhaps…. miracles do happen.
Yahoo has revealed a second separate breach that involves over 1,000,000,000 accounts… that’s right, a one followed by nine zeros! The breach of security has lead to the release of most of their user’s names, telephone numbers, dates of birth and other personal information. As all Flickr accounts are tied to Yahoo accounts the breach affects many Flickr users as well.
If you aren’t going to delete your account then at least change your password. It is likely Yahoo and Flickr will force you to do this anyway.
The following two embeds use the provided embed codes from each site. Literally select the the code for the size you want to display then copy and paste into the blog post.Embed from Getty Images
Compare to Flickr’s embed code:
None of Flickr’s embed features work with WordPress. The reason for this is that Getty’s code seems tailored for WordPress whereas Flickr’s is not.
I have noted an odd issue with the way autotags appear to affect search results.
Many autotags have a hierarchal structure. For example to get the TULIP autotag you first have to get FLOWER and to get FLOWER you have to first get PLANT. Similarly, to get the POODLE autotag you first have to get the DOG autotag and prior to that you need ANIMAL.
When you do a text search on ANIMAL or DOG the relevant results typically offer images with high social activity. However, when searching for POODLE, the relevant results offer mainly older images with low social activity. This pattern is noted with PLANT, FLOWER and TULIP as well, as it is with many of these autotag groupings.
Why would high social activity make animal and dogs relevant but low social activity make poodles relevant? Having different algorithms makes perfect sense but not within one hierarchal structure.
So what? For those who search optimise for relevant, it makes no sense to optimise for 3rd level autotags like tulip and poodle, as the standard optimise techniques do not work for the 3rd order tags. Focus on the more popular 2rd order tags.