I started an experiment. I wanted to compare the two new federated alternatives to instagram when it comes to image sharing. Contender one is the fediverse candidate Pixelfed. Contender two is Bluesky with the new view on images called Flashes.

The experiment was not really scientifically correct green fieldish. I used my Pixelfed account that was about three years old already, but rather small. I started with 44 followers. And on the other side I used my default Bluesky account which also was not heavily used before coming in at 30 followers. Although it looks more neutral, I have to add that I also interact with other types of content with this account, which could lead to some additional follows through unrelated traffic.

So take the results and findings with a grain of salt. As usual, the experiment was mainly for mysef to find out, which platform to pick for myself. And I myself don't have the ambition to become famous with my photos. Everything is more like an observation rather than classic measurements. But I found the results interesting anyway.

The experiment itself is easy to explain. I regularly posted photos of my back catalog from the areas Architecture, Landscape and Macro covering significantly different varieties of styles.

10 images in sum, all of them properly marked up with suitable hashtags (~10 per image) and alt texts (created by ChatGPT as I’m lazy) - identical on both platforms, of course. And here they are for reference:

After posting, I interacted with the 5 most specific hashtags. I looked at 5 other images under that and liked them to promote a visit on my profile. That’s 5 x 5 or 25 likes per post. During that time I explicitly did not start to follow any other artists myself.

Findings: Usability

The clear winner is Flashes/Bluesky.

Sometimes I only saw a blurred mess instead of images in Pixelfed. That ecosystem clearly has a problem with caching/performance and probably has no idea what a CDN is. Downside for Flashes is that there is only an app and not a web application, so on the desktop, I had to use Bluesky itself, which does not offer a specialised browsing experience for photography.

Overall I wish, the clients would distinguish themselves more from Instagram and would use more bold and unique UX ideas. How about a full screen swiping experience, for example?

Findings: Reach

Here, the winner appears to be Pixelfed.

In average, the images achieved 50% more likes than on Bluesky. 5 received more likes, three were roughly on par and only 2 got more likes on Bluesky (so 5:4:2). Two Architecture photos on Pixelfed went “viral” with 10+ reposts. Generally reposting is the essence of reach and it’s not done very often with images.

Therefore the development over time is also not really given on both platforms. You post for the moment. After the image shows up on the feed for the first - let’s say - 7 days, the likes are increasing. Later on they stagnate. When an image is re-shared, it appears again in some feeds. Bluesky has next to zero re-shares, which explains the lower reach. But at least the retention on Bluesky seems a bit better, the reactions lasted longer.

Findings: Image Quality

Instagram’s actual content is flooded with trash in the meantime. Additionally, ads are all over the place and video dominates. The good thing about the federated alternatives is that there are no influencers that want to sell you something yet. And especially there are no ads. Plus points for both platforms.

I like to skim through cool photography, travel impressions from all over the world, phantastic architecture and art pieces as well as stunning street photography. And on top probably a handful of memes and pet photos. That would be my wishlist for a feed.

In order to check that, I conducted a second experiment. I voted 5 x 30 images on each platform on a scale from 1 to 10. 5 times means on fice different days during the last weeks. For Pixelfed, I went through the default home feed and for Bluesky, I used the Photography feed, because of the multi-purpose character of the platform.

In that case Bluseky turned out to be the winner with an average of 5.2 vs 4.4

Findings: Region

Bluesky in general has a strong American bias

Appendix: Posted Images