Examples of (a) a raw image and (b) an engaging filter that adds contrast and warmth, and (c) a less-engaging filter, which introduces artifacts and adds a cooler temperature. Photos: CC-BY-ND ayman on Flickr
By Saeideh Bakhshi and David A. Shamma
Filters: Practically every modern camera-phone app offers them, from Flickr to Instagram to the iOS built-in camera. You can snap a photo and with a tap make it look reddish, or black and white, or faded and aged. It is now commonplace to capture a photo and edit it on the same device, then instantly share that photo with family, friends, or the world. But despite their present-day ubiquity, filters should not be taken for granted; they affect how people engage with photos in significant ways.
In a relatively short amount of time, filter usage has changed from everyday social photographers to serious photography enthusiasts. Our research team wondered why this was the case and if filtering photos changes how people like and comment on them on social media. In answering these questions, we first needed to understand people’s motivations and perceptions regarding filter use. We did this through several semi-structured interviews with Flickr members of various photographic expertise. Then we wanted to to understand how filters affect photo engagement, specifically social engagement such as favorites/likes, comments, and views.
We found that photo enthusiasts on Flickr, despite having access to high-end cameras, still use their mobile devices to take photos and filter their photos with the app. They mainly use the filters to correct errors or improve aesthetics of their photowork. Conversely, many casual social photographers use their mobile cameras for simpler daily documentations like taking photos of things, events, and people. Social photographers often share their photos with family and friends and use and enjoy filters as a method of photo personalization (making the photo unique to them), which they find fun to use and without the need to learn and use a separate editing suite (professional or otherwise). Interestingly, they often look for filters that highlight salient objects in the photos and try to apply aesthetic effects, such as adding color saturation or making the photo look vintage.
Looking at 7.6 million public Flickr app photos modeled in a negative binomial regression, we found that filters boost engagement on the site. Filtered photos are 21% more likely to be viewed and 45% more likely to be commented on. However, not all filters affect engagement equally. Filters that increase contrast and correct exposure can help a photo’s engagement, and filters that create a warmer color temperature are more engaging than those with cooler color effects. More details on filters, photo content, and engagement can be found in our research paper, "Why We Filter Our Photos and How It Impacts Engagement,” co-authored with Lyndon Kennedy at Yahoo Labs/Flickr and Eric Gilbert at Georgia Tech, to be published next week in the proceedings of the 9th International Conference on Weblogs and Social Media (ICWSM).
What we have begun to unravel with this research is an understanding of community engagement through the modern practice of photography. This understanding applies to those creating the photographs as well as to spectators’ engagement with the photos. Our research aimed to find the tacit relationship between content, composition, and social engagement, and included qualitative and quantitative findings. We found that serious hobbyists apply filters to correct their photos, expose certain objects, or manipulate certain colors. More casual photographers like to add artificial vintage effects to their photos and make them more playful and unique. Finally, we found that all these filtered photos are more likely to be viewed and commented on by the community.