500px’s popular section is the holy grail for all 500px users. It’s the section every one wants to be featured in. Like every other 500px user (I guess) I have impressions and pre-conceptions about what makes it to the Popular section. I hate to rely on pre-conceptions however, so I started to wonder about how I could challenge them with available data.
In order to do that, today I spent a few hours tracking the content in the first pages of popular. I looked at 100 photos in the first 5 pages and tracked available data about the photos themselves, and about the users who posted these photos. The idea was to get a better sense of what’s in popular factually in the hope that it can enlighten the question: how do you make it there?
Before I show you the results, I want to explain a little bit more about how I’ve built the data set and introduce a number of caveats. This may seem boring to you, but analyzing data is part of my day job, and I can’t do it without being clear what the limitations are. Please bear with me.
The data set is very simple. For each of the 100 top photos in the Popular section (starting from the top), I tracked the following elements:
- Factual categorization about the form of the photo: macro, hdr, long exposure, etc.
- Factual categorization about the content: portrait, landscape, woman, street, etc.
- Available data about the photo’s user: number of photos, affection, favorites, number of days since the first photo was posted, awesome or not, country of origin, etc.
I then ran the set through SPSS, a statistical analysis tools to produce the results.
- A sample of 100 is too small to do proper in-depth analysis and assume it’s representative. I don’t pretend what I will expose below is representative of what’s in popular every day, but I believe it’s a first step for me to test my perceptions against the trends I see.
- Assessing popular is like shooting a moving target. People are voting all the time, photos are going up and down. In the first three pages, photos were pretty stable, but past the 60th shot, it starts to move. I ensured I didn’t double count any photo, but I can’t affirm that the snapshot is truly the top 100 photos in Popular at the time I did it,
- The title of my blog is despicable pandering, really. No one can tell you how you can get into Popular just by looking at what’s into Popular. In order to accurately assess that, I would need to see what gets into Popular and compare it to what doesn’t get into Popular. If anyone (including 500px) can produce the right data set for me to do that analysis, I will do it gladly, but this is not it, sadly.
With all that behind us let’s move to what you’ve all been waiting for… THE RESULTS!
Form and content:
The first analysis concerns the form and content of the photos themselves. The assumption here is that photos make it to Popular on the sole basis of what they represent. Note that I cannot objectively assess quality, so there’s no quality assessment here. You can see the full results of the Form and Content analysis in Figure 1 at the end of this post. The main findings are as follows:
- 35% of the photos in the top 100 of Popular represent women. Of these, 69% can be described as fashion shots, 29% are nude or sexy, and 2% are travel shots. Also worth noting is that all the nude shots are of women.
- 27% of the photos are landscapes, and 37% of these landscapes are long exposure shots with water.
- 14% of the photos represent plants (flowers, trees or mushrooms) and 11% represent animals.
- 10% of the photos represent cities or architecture. 40% of those are black and white.
- Overall 10% of the photos are black & white.
- Many topics and themes are underepresented (and many are not represented at all). Surprising (to me) were street photography (4%, and I took a pretty broad definition), travel (2%) and sports (1%)
By and large, this comforted me in some of my pre-conceptions: I had already felt that glamour shots and long exposure watery landscapes were predominant, although I thought there would be a higher proportion of nude shots. By and large though, we have a first answer to my title question: if you want to be featured in Popular, shooting good looking women probably maximizes your chances.
Popular User Profiling:
The other thing I wanted to investigate was if certain users were more likely to be featured in Popular than others. Let’s be clear, I can’t answer that with the dataset I have. My hypotheses however were that users who had been around for a long time were more likely to get in, and that users who have had previous photos in Popular were more likely to get in again. I can’t really test the latter. One thing that came to me while analyzing the data was that since nude shots seemed to be “attractive” to viewers, maybe people who had previously posted nude shots were more likely to be in Popular, so I tracked that as well. Finally, I wanted to address the 500px Russian connection allegations.
Here are the main results to consider:
- Users with photos in Popular have an average of 69 photos posted, 1599 photos in favorite and 3860 in affection. Obviously, it’s hard to draw conclusions from these high figures since being featured in Popular not only requires photos with high affection but also generates affection and favorites. Still, the averages suggest that at the very least many of the users with photos in popular are regulars there.
- Having said that, the average “age” of users with photos in Popular is 120 days, 4 months. So it’s not like the Popular section is trusted only by “old” users who have built following and affection over time. In fact, 25% of the photos in Popular have been posted by users who have been on 500px 11 days or less, so the promise that even newbies can make it seems to be largely verified.
- When it comes to countries, 16% of the photos in Popular are posted by users from the USA, as opposed to 11% from Russia (see Figure 2). In one sense, that shows that the section isn’t dominated by Russians as is often thought and said (I plead guilty). Still, one could argue that it’s a surprising proportion of Russians (and more generally Eastern Europeans) compared to an expected profile. Ultimately though, what I found great was the confirmation that many countries are represented, including some you wouldn’t expect to see (more photos from Iran than from France!)
- Finally, having posted nude pics does seem to be a good predictor for being featured on Popular: 27% of the photos posted in Popular are posted by users who have previously posted nude pics (even though only 10% of the photos are nudes).
So, with tongue firmly in cheek, your best chance to be featured on 500px is to be American or Russian, and have posted nude pics in the past.
I hope you found this cursory analysis interesting. If anyone has a brilliant idea as to how I could build a more solid data set, let me know !