Stock Photos

In daily life, you’re surrounded by perfectly generic images. If you’re messaging with a customer service chatbot on the internet, you might be greeted by a photo of a smiling woman who’s wearing a headset. If you’re flipping through a brochure of investment options at a bank, you’ll see a shot of a diverse group of people in suits in a brightly-lit conference room. And if you pass a bus stop advertisement for a local gym, you might catch a glimpse of a runner racing off into the sunset.

ARCURS: Our stuff is the secret background of the whole world. 

That’s Yuri Arcurs. He’s the C.E.O. and founder of a company called PeopleImages. And he’s one of the most prolific stock photographers in the world.

ARCURS: Essentially any big city, if you walk for ten minutes, you’ll probably see five or six of my shots. Pretty much all the big brands at some point or another have used our files.

When a company needs an image that embodies a particular vibe in a bland but professional way, it often buys a stock image — an existing photograph that can be licensed for commercial use. There are millions of stock photos for sale on websites like Shutterstock and Adobe Stock, depicting nearly every subject you can dream up. If you need a photo of an elderly Japanese woman on a tractor, or a heavyset man in a blue shirt eating a donut, a stock photographer like Arcurs probably has you covered.

ARCURS: The total database we have right now is 600,000 images.

But with commissions shrinking and new kinds of competition on the rise, it’s getting harder and harder to make money in the stock photo game.

ARCURS: It’s got to be the right balance. It’s got to be well lit. It’s got to have good composition. And even if you have all those elements put together, you’re probably going to get sales for about a year and then it’s over. 

For the Freakonomics Radio Network, this is The Economics of Everyday Things. I’m Zachary Crockett. Today: Stock photos.

*      *      *

120 years ago, if a newspaper or magazine needed a generic photo — say, a shot of the White House, or a family enjoying a picnic — it would have to send a photographer. But in the mid-20th century, a number of companies stepped in to let them outsource that service. They assembled archives of photos with broad applications and provided them for a fee. This model became the norm for journalism, advertising, corporate communications, and other businesses in need of basic images. Back then, the ordering process was a bit cumbersome.

ARCURS: They sent out these big catalogs where you had to go page by page and choose your file number. You would call and say, “I want this file number, XY2467.” And then you would be sent a film with that file that you could then scan and put in your print.

In the 1990s, archives of physical photos were transferred onto computer servers, and stock agencies sold CDs with preloaded images. There was a limited amount of space available on these CDs. And the photographers who made the cut were handsomely rewarded.

ARCURS: The photographers that were in this little elite clan, they were making a massive amount of money. They were selling at very high prices. And then on top of that it would be “rights managed,” which means that if you wanted to print it just a tiny bit more than what you said you wanted it in, you pay extra. Or you put it on a brochure, then you pay extra. So it’s very tightly managed, and incredibly tedious for the consumer.

But in the 2000s, stock photography experienced a revolution. Websites started popping up that allowed anyone to upload and sell their photographs.

ARCURS: Customers started choosing cheaper files with less restrictions They might be paying a dollar per download, versus $1,200 or $600.

This new model was called microstock. And it opened the door for amateur photographers like Yuri Arcurs. While he was a student in Denmark, in 2006, he started posting his photos to one of the bigger microstock websites, Shutterstock.

ARCURS: I was studying psychology, so I needed the money. I was dirt poor. I decided, “Okay, I’m going to put three months of my time just into shooting as much as I possibly can.”

Good timing, an eye for the right images, and a lot of work led to quick success.

ARCURS: I would see my shots in the street or on newspapers and things like that, I was ecstatic. And then it went so fast. I went from making $500 one month, $4,000 the next month, $12,000 next month. I remember having gone from having a hobby to making about $40,000 a month. That was about two years in. And at that point I had some pretty tough decisions, because this was obviously more money than I was going to make as a psychologist. I’d have to be Dr. Phil to make that kind of money.

Today, Arcurs runs a 20,000-square-foot production studio in Cape Town, South Africa. His company, PeopleImages, does 3 to 4 shoots every day, and employs more than 100 people — photographers, stylists, editors, lighting crewmembers. And he’s amassed an archive of more than 600,000 stock photos and 250,000 stock videos, which he sells on many different online platforms.

ARCURS: Adobe Stock, Shutterstock, Dreamstime, 123 RF, FreePik, iStock, Depositphoto.

These sites host hundreds of millions of stock photos across nearly every category imaginable. There are doctors gleefully showing charts to patients, close-ups of men in suits shaking hands, and dogs lounging in hammocks. Search for the word “marriage” on Shutterstock, and you’ll get 3.4 million results. “Sunset”? More than 14 million.

For Shutterstock, the business model is fairly simple: Customers pay a monthly subscription fee for a set number of downloads, or they can buy images individually. A standard license grants buyers broad rights to use the photo anywhere in the world, forever. For the photographer, the payout comes in the form of a royalty — a percentage of the sale price for each photo. That payout ranges from 15 percent to 40 percent, depending on how many photos the photographer sells in a given year.

ARCURS: They’ve been fairly successful, all the agencies, in implementing all kinds of tier systems and brackets and levels to try to not pay out as much as they need to. If you don’t keep feeding the beast and upload all the time it’s very hard to get up those ladders, so to speak, and get the higher royalties. And that can take you five years to get all the way up to the top of that — until then the sales will be very small.

Most stock photographers earn around 30 cents every time someone licenses one of their images. Arcurs, who is at the top of the food chain, says his average payout works out to around $1. And on a typical day, he’ll log more than 10,000 sales. While Arcurs has more than half a million photos, only a few of them attract a lot of buyers.

ARCURS: About 10 percent of the whole portfolio will be 90 percent of the sales. It’s a very congested market with a lot of competition. A lot of files will be on a stock site and circulated for a couple of weeks. And then the algorithms will determine that it hasn’t got enough, you know, markers in terms of clicks, views and downloads, to keep being surfaced and then they’ll just dump it in the background.

One financial measure of a stock photographer is return per image, or R.P.I. It’s how much revenue each image generates every month, averaged across the photographer’s entire portfolio. It includes the thousands of photos that don’t sell at all.

ARCURS: An R.P.I. gives you a rough indicator of how successful your assets are. So if your R.P.I. sits around $1 per asset per month, then you have an incredibly high R.P.I. I think the average will be cents to the dollar.

If you want to be a successful stock photographer, you can’t just upload random shots of your grandma eating a hamburger. You have to know what’s going to sell.

ARCURS: You can look at what’s popular on sites. The only problem with doing that is that you’re essentially replicating something that’s already there. So that’s a source of inspiration you want to minimize. 

The better approach, says Arcurs, is to try to get ahead of your competitors by keeping a constant eye on trends.

ARCURS: You’ve got to be good at reading between the lines a little bit in terms of what’s happening in the media. Elon Musk is shaking hands with Trump. Okay, noted — politics. Let’s shoot that. And then now you just need to put your creative mind to work and start to imagine, “Okay, if I’m going to write an article about politics, what kind of pictures do I need? Okay, I might need a picture of people putting in their ballots in the voting station. Okay, I’m going to take that shot. They might be in doubt. Okay, I’m going to take that shot. They might be convinced by their friends. Okay. I’m going to try to stage a scene that looks like that.”

The artistry of a good stock photo lies in knowing how to visualize an abstract concept.

ARCURS: Let’s say we’re focusing on making a factory. Ask yourself: what kind of shots would the owner of the factory want? Well, he would probably want his staff lined up outside the factory with the factory in the background, and they look confident and they’re dressed nice. All right, so we’ll shoot that. You’ll probably want some detail shots — some hands, some fingers doing some really intricate and difficult things. He would want to show off the machinery.

But stock photos need to have many applications. Arcurs’s hottest-selling asset at the moment, for instance, is a photo of a woman taking a break from a run, with her hands on her knees. She’s standing in front of a bright blue sky, smiling, and gazing hopefully toward the horizon.

ARCURS: That can be used for so many things. It’s kind of uplifting. It’s got positivity. It’s a picture that could be used in a commercial for a bank, but also an insurance policy or, you know, want to take some time off or get in shape. So it’s got all these utilizations open. That’s really good for sales.

Photos like this look pretty simple on the surface. But creating them requires pricy photoshoots, models with a specific look, and some technical wizardry.

ARCURS: It is not like the good old days where you could just take a bunch of young people, put them in a suit and put them around the table and off you go. Those days are over.

That’s coming up.

*      *      *

Once Yuri Arcurs settles on a concept for a stock photography series, he and his team usually spend at least two days researching the topic. Arcurs takes realism very seriously. If he’s producing a series of shots of a computer engineer at a desk, he’ll make sure that real code is displayed on the monitor. If he’s staging a rock concert, he’ll hire a band to play on stage and put on a real show. And if it’s an office-themed shoot, he’ll ask the models to give actual Powerpoint presentations.

ARCURS: Obviously it is staged, but it would be very difficult to pinpoint that this is staged. You want to catch the model in the middle of a sentence actually talking. So when you direct a set like this, you want to give them talking points. You want to actually have them talk about what’s on whatever display media they’re using, and make sure that they don’t just say nonsense, they actually try to explain stuff.

On the day of the shoot, there is a strict plan for every shot — because wasted time is expensive.

ARCURS: You need a location, you need a fairly big crew, because you’re going to be setting up in the morning. You need lights, you need catering. You need four or five, six models. You need at least one producer, probably a focus puller, camera operator.

A recent office-themed shoot Arcurs did, for instance, cost around $5,000 and yielded 106 images. That’s an investment that won’t immediately be earned back.

ARCURS: We look at how much a shoot will bring in in its whole lifetime and have a rough idea about when it’s going to break even. Generally, the break even point for shoots is about two to three years, which unfortunately means that if you want to start out in this industry, you need a pretty big loan.

Another challenge is finding the perfect model. Traditional modeling agencies usually can’t give Arcurs what he’s looking for.

ARCURS: The model type that they have is a very, like, high fashion model type, runway model look, we can’t use that at all. For us, a skinny person is a problem. It is not a good asset to have on one of our shoots, because it stands out. There is a general misconception that prettier people sell better, but that’s actually not the case. The case is that people that look trustworthy sell well.

Arcurs looks for people who pass what he calls “the weekend test.”

ARCURS: The weekend test is essentially just that you look at a person and, just judging from their looks, you have to now decide really quickly if you would want to spend a whole weekend on a stranded island with that person. It’s the guy next door, girl next door. They have some appeal. And they look just approachable.

Because stock photos attempt to replicate real life, there’s a demand for people of all ethnicities, races, sizes, shapes, and ages.

ARCURS: Some of the best selling models are older people with good faces. Senior models sell extremely well and are very willing to shoot.

To find models, Arcurs puts out casting calls on social media. He puts a sign out in front of his studio in Cape Town for walk-ins, and will sometimes stop people on the street who he thinks have the right look.

ARCURS: One of the best places to get a really good, diverse group of models is farmers’ markets. They’re amazing, really stylish, trendy, healthy people.

Stock photography models don’t get royalties for the photos they’re featured in. They’re typically paid a day rate. But sometimes, they get another payoff in the form of celebrity.

ARCURS: You can get a lot of funny fame stories from the models that have been circulated the most. One of our top selling guys is called Rodrigo. He’s everywhere in Brazil, and they love him for it, so he’s become this little local celebrity. If you look at the same pretty face or a handsome guy for ten years straight, it almost becomes a personal relationship, I guess.

Photos purchased through microstock agencies like Shutterstock can be used for any purpose, so long as it’s not used to promote Illegal, pornographic, malicious, or defamatory material. So stock photography models never really know where they’ll end up. They might be on a billboard in Slovenia, or in an internet ad for a senior home in Kentucky. Sometimes, this can lead to problems. Years ago, a government agency in Germany purchased one of Arcurs’s photos and used it in a campaign for H.I.V. awareness.

ARCURS: It was actually one of our skinnier models, one of our guys, who just in one shot happened to look a little, little bit sickly, if you will. And then they photoshopped it a little bit, and they made it look even worse. We have had many cases where models have come back and said, “Hey, what is this? I didn’t agree to this.” And then we have to chase down, whoever did it and get them to take it down. Sometimes it requires lawyers, letters. It’s not the most pleasant side of the industry.

In a few cases, Arcurs’s photos have gone viral online. In 2009, he took a shot of a young woman with her head in the palm of her hand, smiling into the camera. On Reddit, it became the template for Good Girl Gina — a meme often overlaid with text about what a great girlfriend she is. The model, who goes by Emma Katherine, is now an internet legend.

ARCURS: She got very famous from that one shot. So yeah, she’s been very happy about it.

The image has been used thousands of times — and almost none of those meme-makers bought a license, which means that what they’re doing could be a copyright violation. But Arcurs says it’s not worth the trouble of pursuing legal action with anonymous memers on the internet.

ARCURS: You can forget about trying to monetize that. That’s not going to happen. You’re going to have to just laugh. But, I kind of enjoy it, to be honest. I’ll forfeit the royalties for the greater good of the meme world.

Now, most stock photos don’t get turned into memes — which means Arcurs has to do some work to get them noticed. When a buyer searches for, say, an image of a flower in a field, they’ll find hundreds of thousands of options on Shutterstock or Adobe. One of the most important steps for a stock photographer is indexing each photo: adding keywords that give their work a better chance of showing up on the first few pages of search results. Arcurs employs a team of indexers. They attach extremely specific — and sometimes, abstract — keywords to every image.

ARCURS: It’s a very nerdy enterprise to be, keyword person. You’ve got to have this enormous vast amount of common knowledge, spell perfectly, and then know all kinds of strange words and weird things.

CROCKETT: What kind of weird keywords are we talking?

ARCURS: I mean, one picture can be of a certain flower taken in some place in the world, and now we have to download an app and scan that flower, find out what the real name is, where it is, and then suddenly you find yourself looking into botany for two hours. And then the next picture might be a rifle, a hunting shoot. So now you have to dive into weapons. And is it loaded or not, magazines, and distance, and scopes, and position they’re sitting in. If you do common and really high ranking keywords all the time, you’re not going to sell because you’re competing against too many files. So the more specific your keywords are, you have a way better surfacing chance.

Even with the best concepts, models, and keywords, the business of stock photography is getting tougher. There are now websites that offer stock photos for free. And they’re operating on an entirely different business model than sites like Shutterstock.

ARCURS: The business model is to disrupt enough and be enough of an annoyance that somebody is going to have to buy you out. What they’re really good at is making two or three pages of content look nice. But the rest of the portfolio behind it is actually pretty junkish.

More recently, stock photographers are under threat from machine learning technology. Tools like DALL-E and Midjourney allow users to type in a prompt and generate increasingly realistic images of any subject. NVIDIA, OpenAI, and Meta all have partnerships in place that allow them to use stock photo libraries to hone their machine learning capabilities, or generate images. But, at least for now, Arcurs still thinks he and his human peers have the upper hand.

ARCURS: A.I. is definitely a big shakeup. Is it a threat for the industry? Yes and no. It is a lot easier still to buy on a stock site than to start writing prompts and getting something to really generate the shot you’re after. You’re looking at probably 2 or 300 generations to get one shot, and that’s about 2.5 hours of work to get to that. Let’s say you want to change something. You know, I want this hand that’s shown to the right side to just be a little bit more to the left, with some copy space above it. You can write all the prompts in the world you want, but you are not going to get the A.I. to understand that. So these kinds of smaller changes that a professional needs are very hard for an A.I. to execute. For now, it’s just not refined enough to satisfy a really finicky art designer out there. And we haven’t gone down in income because of it.

And, there’s one more thing that an algorithm cannot do: partake in the joy of seeing its handiwork out in the wild.

ARCURS: I used to tell people, “I took that shot!” They would just look at me like I was some kind of Looney Tooney. Like, “Okay, there’s a screw loose here.” It can be pretty hilarious sometimes.

*      *      *

For The Economics of Everyday Things, I’m Zachary Crockett. This episode was produced by me and Sarah Lilley and mixed by Jeremy Johnston. We had help from Daniel Moritz-Rabson. Also: huge shout out to our listeners, Kally Kumnick, and her boyfriend, Dylan James. Kally made her own special episode of the Economics of Everyday Things and gave it to Dylan for his 24th birthday. If you’ve got something cool like this to share, or if you have a suggestion for an episode, feel free to email us at everydaythings@freakonomics.com. All right, until next week.

ARCURS: Keywording — yeah, if you’ve done that for three years, you are the person nobody wants to play against in Trivial Pursuit.