Photographing Shelter Dogs: You (Need to) Let the Dogs Out!

About 670,000 shelter dogs – almost half of the population – are euthanized each year in the United States (Source: ASPCA). Approximately 80% of cats and dogs euthanized in shelters each year, however, are healthy and adoptable pets (Source: Humane Society).

Websites like Adopt a Pet and Petfinder have become a popular way for potential adopters to look at dogs in shelters near them. With modern society revolving around use of the Internet, websites like these use photos of dogs to draw in viewers. Many photos of dogs are presented in a way almost like clothing in an online store, and when people see a dog they find physically appealing, they can click on the photo for more information.

Therefore, a dog’s photo can be extremely important in determining whether they will be adopted. Photos of dogs online vary drastically, however, ranging from blurry and partially obscured images of unkempt pups to high quality photos of dogs in adorable poses. To try to gauge what the best approach is, I conducted three experiments to investigate the effects of altering certain aspects of shelter dogs’ photographs on people’s perceptions of the dogs’ adoptability.

Before I began, a quick internet search revealed a wealth of opinions on how best to photograph shelter animals. While some writers shared their opinions on the best backgrounds on which to photograph dogs, others gave advice on how zoomed in a photo should be, the best lighting, and even about the use of funny hats and costumes. Interestingly, some of the advice was contradictory. For instance, some articles suggested photographing dogs on a plain background to make them stand out, while others recommended integrating the dog into a busy scene of toys, attractive backgrounds and human companions. People were happy to share opinions, but almost none of their advice was supported by actual research.

One of the few actual studies showed that a large, potentially aggressive looking dog could be helped by being photographed with people. In one study from 2015, it was found that a pit-bull-type dog photographed with a boy and an older woman was perceived more attractive than the same dog photographed alone, and that the addition of a “rough” man actually decreased the dog’s perceived attractiveness.

For my study, I replicated the experiment using a different dog, a boxer, and tried to add a greater degree of control by standardizing the position of the dog. In all three experiments, participants were recruited through the website Amazon Mechanical Turk, where they were given a link to the survey on SurveyMonkey. In Experiment 1, 120 participants were randomly assigned to one of four photos where I paired the boxer with a (not particularly rough-looking) man, a similar looking young boy (his son), and the two of them together to see what the cumulative effect would be. Interestingly, the dog’s attractiveness ratings were equivalent across all four conditions, demonstrating that the inclusion of people in photographs of shelter dogs won’t necessarily have an effect.

Curious to explore other factors that might make these shelter dogs more appealing to potential adopters, I continued to look into the various strategies that were suggested online. A common topic was the idea that a dog should be photographed with signs of personality such as through the use of funny costumes or even something as simple as a head tilt. In Experiment 2, I recruited 200 participants to test whether a dog with a tilted head or visible tongue would be seen as more attractive than the same dog with its head upright and tongue in its mouth. Given that dogs often cock their heads and loll their tongues, these seemed like easy, natural photos to take; unfortunately, neither head tilt or tongue visibility had an impact on perceptions of shelter dog attractiveness.

Another seemingly simple way to manipulate photos was through framing or zoom. A few articles recommended zooming in on the dog’s face, allowing it to take up as much of the frame as possible and therefore causing it to be the center of attention. Following this advice, I compared the attractiveness ratings 220 participants had given to the same photo zoomed in and zoomed out. Again: no difference.

In the same experiment, I also tested what, during my search for a companion animal, had struck me as a terrible way to portray potential pets: in cages. While I had read recommendations about what to include or not to include along with a dog in a photo, I had never seen mention of a cage.

The use of cages reminded me of the commercials that show picture after picture of despondent animals in need of care while a voiceover describes the urgent need to help. These commercials always make me turn away or even change the channel; I find them too distressing to watch. In addition to discouraging people from looking at the photo due to sadness, I thought a cage could make a dog appear aggressive.

I decided to compare a dog on three backgrounds: a sunny outdoor setting, a plain, indoor background, and of course, the infamous cage. My hypothesis was that the outdoor setting would create the most positive mood and therefore result in the highest attractiveness ratings, and that the cage would have the opposite effect. While the cage did depress the ratings, the dog photographed on the plain indoor background had the highest ratings, suggesting that there may be benefits from allowing a shelter dog to stand out from a simple background.

While these experiments provide some insight into the world of adoption photos, the information they give shows us that the specifics of shelter dog photography may not be as important as many people seem to suggest. That’s not to say that shelters shouldn’t try to portray dogs in an appealing manner, but it’s probably unnecessary to hire a professional photographer or try to adhere to the plethora of suggestions found online. It could be that more important than improving the quality of shelter dog photos is figuring out how to get more people to view them; after all, love at first sight can’t occur without that first glimpse.