I’m a nerd. There, I said it. It’s out there now, and it’s never coming back. I’m adamant that all facets of life are infinitely improved by statistics. I paw over numbers, percentages, and graphs for academia, sports, science, films... the list rolls on. Even reeling off the sort of stats I like makes me want to forge some sort of Excel spreadsheet to identify the stats for which areas benefit the most from stats. Sorry, I digress. The point is fewer things are richer in information than statistics. We often use this approach to compare lenses and cameras, but what if we could apply it to something far more subjective: portraiture?
Further still, if we could identify themes present in popular portraits and apply them to our own portraiture, would it make our work more popular? Firstly, I’ll qualify what I mean by "portrait," as I noticed in my previous articles that my fellow pedants aren’t at peace with the way in which I am using the word. By "portrait," I simply mean a photograph (in this instance) of a person or persons. I am side-stepping the classic usage where it usually means only the face or head and shoulders.
I’ve mentioned in my previous articles that I have run Acufocal Portrait of the Day for a long while; at the time of writing this, there have been 836 portraits of the day, which are only selected Monday-Friday to provide 20 per month. That means my Pinterest board has over three years of portraits pinned to it. This recently got my number-loving cogs whirring, and so, a mini-project was born. My love for statistics encouraged me to keep careful analytics on everything I can and I decided to see what I could learn about portraiture from the varying interactions with these portraits of the day.
I didn’t create my Acufocal Portrait of the Day board with this in mind, so I can’t pretend this is highly scientific and has unwavering accuracy. What I can do, however, is identify trends and see what makes a portrait great, or if not great, popular. And let’s face it, we’re in the era of social media where popularity is just a synonym for great. That said, popularity and greatness have essentially always been the same thing. Alexander the Great was only great because he had so many followers, albeit military as opposed to Twitter. The more things change, the more they stay the same. Anyway, enough about Alexander the Popular.
All 836 portraits of the day I have selected have been from 500px, and that’s useful: I will limit my analysis to just portraits on that site. I will make a separate Pinterest board of the top 50 most popular portraits based on interactions and impressions and then try to identify commonalities between these images. Although this isn’t an infallible method by any stretch of the imagination, it does have a few perks over just searching for the top 50 most popular portraits on 500px. For example, the photographer’s following doesn’t play such a huge role; when photographers of Sean Archer’s celebrity upload an image, they will get an enormous number of views, likes, and shares. I’m not debating whether Sean deserves this reception (he does), but it’s not helpful in this instance, as a photographer with ten followers uploading a picture that gets half Sean’s views would – in theory – be of higher merit.
There are a few problems with this method that are worth noting. Problem one is that the longer the portrait has been pinned, the longer it has had to receive impressions and interactions. There isn’t a lot I can do about this, but it’s worth noting that some of the top 50 are more recent than you might expect (as recent as this month), which could be due to problem two: increased following. Over the years, this board and the places I post it have grown in popularity, albeit not explosively. This means that newer pins tend to receive more impressions and interactions than older pins did when they were first posted. I’m hoping that problem one and problem two cancel themselves out, even if only partially. There are numerous additional small problems, like certain days and times of posts generating more traffic than others, but this can’t be helped. I’m not looking to provide a definitive and conclusive answer to the question, but rather explore trends in popular portraits. Enough of the boring stuff, let’s get to looking at pretty pictures and what they have in common.
48 of the 50 images are women and two are young boys. This is an unfortunate bias that has been around a long while. I became conscious of this in my Portrait of the Day selections quite early on. Being a straight male, I accused myself of being biased towards my favorite gender to look at, and I plodded down a different and arguably worse road: reverse discrimination. I was trying so hard to even out the gender numbers within my selections that I began picking images I liked less merely because they were of a male model rather than a female. I corrected this soon after, but unfortunately, 500px is populated with primarily images of females rather than males.
It’s difficult to say what exactly a bright color is, but I’m defining it loosely as a vibrant color that stands out and is a key feature of the image, that is, more than simply luminous eyes. I would say it’s around 15 of 50 photographs that feature vibrant colors as either a large splash on or around the subject or as the background.
Black and White
Black and white portraiture is usually seen as classier than the Queen being fed grapes in the back of a Bentley, and so, I thought it would feature prominently in the top 50. To my surprise, black and white images only account for 10 percent of the selection. That said, there are significantly more color images than black and white both in my 800+ Portraits of the Day and on 500px in general. Nevertheless, I feel they have underperformed, and I wonder if their lackluster popularity is more a sign of the times with post-processing being commonplace nowadays.
Whether eye contact is necessary for a great portrait is debated to unfathomably boring lengths. What I can say is of the top 50, 74% have eye contact with the camera. That is a pretty hefty bias and one that’s difficult to ignore. Curiously, however, two of the top five best-performing portraits across all analytics didn’t have eye contact. I worry that the reason one of them performs so well isn’t the great lighting and pose, but rather that the only element of the frame in focus is the pair of voluptuous lady-udders that are obvious even at thumbnail size. We should all feel ashamed as a collective.
One thematic observation I made was a more unexpected one for me. A whopping 46% of the images feature women with blue eyes if you include images where the eyes aren’t a dominant feature and one black and white image in which the eyes are still obviously blue or green. It is estimated that only 10% of the world’s population have blue, green, or silver (very light blue) eye colors and that brown and hazel eyes make up in excess of 60%. There are too many considerations to accurately and fairly surmise why this might be the case, but however the penny falls, blue eyes make for engaging portraits, it would seem.
Personally, I love freckles. I’d quite like to punt the shin of whichever dolt at Match.com indirectly labeled freckles as "imperfections" in their last ad campaign on London’s underground. They add so much to an image for me, and I was pleased to see that 14% of the top 50 featured those golden dermatological sprinklings.
I didn’t quite know how to word this, and I’m still not sure. I find there is a slight dichotomy with photographers taking portraits that don’t have an express purpose (i.e selling a fashion garment or a jewelry range, etc.), and although they do often overlap, identifying them creates a platform to explore. The first category is photographers mimicking a look. To an extent, everybody does this, but some more than others. There is nothing wrong with enjoying the way certain images look and striving to create similar photographs in the same vein, but some do it with no interest in standing out. The second camp attempts something unique within their image that will make the result more memorable. It is this second camp I find the most relevant to this mini-project.
A large portion of the top 50 (it would be fruitless of me to try to give an exact percentage here) have something about them you don’t often see. I realize this is frustratingly vague, but there is something to it, whether I can get at it or not. There are several shots that are fairly unremarkable, albeit excellent, but the majority have something unusual. This might be styling, a prop, location, or even as simple as the toning of the colors in post. What I feel I can assert rather confidently is that images that take a step further from just being a nice, perfectly exposed and composed portrait seem to generate the highest levels of clicks and shares.
Mining for Gold
A lot of the information derived from the top 50 is intuitive, that is to say attractive and young female models will usually win out in popularity over portraits of unattractive old men; that news isn’t going to break the internet (reference intended). However, if you’re a photographer who takes correctly exposed and well-composed portraits, but aren’t receiving the recognition your heart (read: ego) desires, then perhaps try the following:
Firstly, a young and attractive female model is going to go a long way if that’s not already the demographic you shoot. I’m not saying that’s right, and I’m not saying that’s all you ought to do, but it clearly works. Direct your anger for this inequality wherever you see fit – me, if need be – I’ll leave my mobile number in the comments.
Secondly, use a model with blue eyes and make sure those beautiful blue lamps are fixed on the lens. My old boss told me once how she traveled to Burma, and as she walked around, people would come up to her just to stare at her blue eyes as they rarely (if ever) saw that eye color. This fixation appears to not be born of rarity but something else if this micro-study proves anything.
Thirdly and finally, escalate the image in some way. It’s almost as if you ought to dream up a lovely portrait: an attractive model, great light, and composition, and then before you shoot, throw something in the mix that will make your image stand out from the crowd. This could be through a milliner’s complicated creation, an unusual prop, or something far more bizarre. Whatever you choose, it appears that the unexpected can even be worth sacrificing some of the "niceness" in your image.
I would love to hear from any photographers who try any advice in this piece. Did you see a spike in views, shares, and ratings? Was this absolute rubbish of no consequence to anyone? (I’m a medium-sized man; I can take it, I think.) Also, study the top 50 yourself, and tell me in the comments if you spot any trends I missed or contest any trends I identified.
Click here to view the full top 50.
Lead image by Dani Diamond and used with permission.
Tell me about it. I hope all red-heads that fall victim to school yard bullying realise how revered they are by photographers!
After reading the comments I noticed a red head with freckles in the "features photos" section :)
If I could give "two thumbs up" to red hair, I would!
100% agree... and an absolutely gorgeous young lady and beautiful portraits by a young photographer here on fStoppers needs recognition:
I had a red headed girlfriend once. She was completely wacky but OMG what a great time. BTW I love to shoot redheads they are much more interesting to look at, freckles and all. Some are stunningly beautiful.
An interesting article on analytics. Some of the information analyzed is objective (meaning that it can be measured) like freckles and eye color; others are subjective (meaning that it is an opinion) like uniqueness.
I'm analytical myself. when deciding what DSLR to buy, I created a spreadsheet of the features my film cameras had (full frame, 6 FPS with the respective motor drives on the A-1 and F-1N) with the DSLR models considered. The closest match in features wanted was the 5D Mk III.
I really like how you have reverse engineered this concept of photo popularity. I think it is very useful information to have. But I do have one problem with the concept. And that is that you are basically following trends. Ok uniqueness is not trend.
This is great for most Photogrphers. But if you really want to build a following it is better to be ahead of trends. So that means making a risky move and going away from what's popular.
It's harder to do but in the end you will see a lot more success from it. Maybe.
I like the idea of this article, but unfortunately I don't think this analysis really gives insight into what makes a portrait popular. Just presenting descriptive statistics within the sample of pre-selected "popular" images isn't the same as saying those variables EXPLAIN the image's popularity.
A better analysis would have accounted for the base rates of those features in the full pool of portraits on 500px. In other words, these percentages might just represent ALL portraits on 500px. For example, the article states: "74% have eye contact with the camera." If that's the same percentage as ALL of the pictures on the site, it means that eye contact has no effect on popularity. Alternatively, if 50% of all portraits have eye contact, then the fact that 74% of the popular ones have eye contact suggests that eye contact is related to popularity. On the other hand, if 90% of all portraits have eye contact, then the fact that only 74% of the "popular" ones have eye contact suggests that eye contact actually HURTS an image's chances of being popular.
So without the context of the base rates (i.e., the probability that ANY portrait has these features), the statistics in this article doesn't necessarily tell us anything about popular portraits specifically.
Nevertheless, I love the article, and it highlights some interesting considerations when it comes to portrait photography. I just think the analysis might be a little misleading.
Thank you for your reply Andrew. I agree and I did say it was far from being conclusive. The problem with using all portraits on 500px is -- as I mentioned -- that the following of the photographer is by far the most powerful influence over a photograph's rating/likes/views. The benefit of a Pinterest board is that a photographer's following doesn't play anywhere near the same role. Admittedly, however, a smaller sample yields less reliable results.
The formula in 3 words: pretty white women.
Just look at the "featured photos & videos" at the bottom of the page - common thread to the constant stream: pretty white women.
Race isn't a correlation I had drawn but it's an undeniable one now you've pointed it out. How very depressing.
Why? If a majority prefer to look at white women, what's wrong with that? Diversity is wonderful but one shouldn't demean the mainstream in it's service. In a society of uniform diversity, nothing would be special.
Something else that strikes me is that very few are 'obviously' studio portraits - most appear to be shot in a natural environment adding an element of background interest. I guess that this is adding to the Uniqueness classification.
Loved the article. Perhaps I should make a spreadsheet to compare your analytics with my own photographs.
Freckles is (are?) always a winner with me, as is red hair. The look of an Irish lass would be very popular.
Thanks Robert, for the great article!
I must say I'm not surprised about b&w (black and white) photos not being too popular. I started using new hashtags on ig for my photos a few weeks back. I saw an insane spike in engagement for my photos with more vibrancy (of the beach). But when it came to b&w (a female portrait), it seriously lacked as much growth in engagement.
I just continue to assume that b&w is for those with the preference for it. Perhaps an acquired taste to photographers and certain other people.
Yet to test out a colored version from the same shoot with the same female subject (with blue eyes).
Actually, it's interesting you should say that. I found the exact same trend in terms of vibrancy/saturation bringing about more engagement on Instagram, EXCEPT with portraits of people. For some reason, my black and white headshots do far better than their coloured counterpart, both male and female subjects.
Awesome article! A great read with a very different approach to improving interactions with images. I'm digging the analytics.
Really curious as to why my comment was deleted regarding you referring to a woman's breasts as "lady udders" and how incredibly misogynistic it is.
Hi Leslie, I haven't seen the comment you're referring to and I get notified when someone comments on my articles. Either way, "lady udders" was merely playful innuendo in lieu of "breasts" as I was making a rather contentious point about why the image is popular and I wanted to keep the tone light. I have neither hate nor disdain or distrust for women, so I'm sorry that's what you have inferred from something meant to be nothing but whimsical.
Calling it "playful innuendo" is the same concept that governs the whole "boys will be boys" movement. You didn't mean anything by it, so it's okay to say it even if it is hurtful? There are MANY different ways to keep the tone light and playful without objectifying the subject of that photo. To be honest, I was enjoying the article until I read that, then I cringed. It wasn't funny. It is an article written by a man, assuming he is talking to a roomful of men (because women aren't photographers, right?) and as soon as a woman says something about it, she has "inferred" something. I, for one, have never thought of udders as whimsical. And the reason the image is popular isn't contentious at all--people like to look at beautiful things (beautiful bodies included) so that's why the image is popular. The only thing contentious about this whole thing is the attitude that using that language is okay.
Look, I'm sure you're a nice guy. And I'm sure to lots of people it sounds like I'm being crazy. But that is the thing about stuff like this-- women are told every day they are overreacting to things that make them feel uncomfortable, mistreated and undervalued. It happens SO often that society thinks it is okay. And all of this adds up, and the collective pain of all of this hinders women in many ways. And if you respect and admire women the way you say you do, then you can either realize that as a woman, I'm saying your words made me feel like that beautiful person was the butt of a joke, and it made me feel uncomfortable. Or, you can drop the term "voluptuous lady udders" to a woman you don't know very well, and watch her reaction. I imagine it will not be positive. It hasn't been amongst my friends as we have been discussing this.
Robert, you laid out the "recipe" for popularity of a portrait. Also found it kinda surprising but true: it is "pretty white women" hmm...