Instagram has evolved immensely since it was founded in October of 2010. The world’s top photo and video sharing mobile app originally launched for the iPhone and geared itself primarily towards a millennial audience. Now the platform functions on a wide variety of mobile operating systems and has expanded to the web. The network today hosts 400 million active users who post an average of 80 million photos per day. Although Instagram is still popular among millennials, nearly a third of adult internet users over 30 also engage with the platform.
It’s not uncommon for communications departments to use Instagram for public outreach and service announcements; however, deeper analytic potentials of the medium are often unappreciated. How can transit agencies go beyond simply posting photos and counting the number of followers they amass? What more can Instagram do for you as a strategic intelligence tool?
Like other social platforms, Instagram does indeed help you communicate; the real advantage for actors in the transportation space, however, is that Instagram offers social media and network data collection benefits that provide unique analytic capabilities. More than just a window into real time conversations, social media networks also function as sources of geospatial information. Each time a user “geotags” a post, that item is assigned longitudinal and latitudinal coordinates that pinpoint exact locations in a city. Geotagged posts are extremely valuable to transit agencies because they allow you to match abstract messages in the digital sphere to concrete sites in and around your transit service. While the geotagging function is available across most social media platforms, Instagram posts are geotagged at a higher rate; it’s estimated that 30% of Instagram posts are geotagged compared to 2% on Twitter, furnishing a comparative wealth of metric data for planners, public relations officials, and other players.
In a pilot project that AlphaVu conducted with a transit agency, we harnessed Instagram’s abundant spatial data by measuring the frequency of Instagram posts within the known radius of a given number of transit stops. In this study we were able to observe which stops experienced higher levels of Instagram engagement. This data also provided an understanding of which stops were most frequented by social media users, and even the demographic makeup of these users. Tracking Instagram data in and around your transportation network isn’t just limited to the demographic variables we explored in this project. Instagram data also lets you tap into the experiential aspects of people interacting directly and even indirectly with your service.
In the pilot project cited above, we observed an Instagram user post a photo of a rail stop with the caption, “Starting to hate train stations”. In another post, a user snapped a photo inside a metro train and pointed out the fact that other passengers were keeping to themselves and coming across as unsociable. Photographic posts like these provide an embedded, situational sense of transit users’ experience: you can discover what your customers are experiencing at the ground level in a way that conveys emotional salience beyond mere text and doesn’t require the physical presence of staff or researchers. The photos illuminate a richer, multi-dimensional experience not as prone to the sarcasm that often pervades other networks.
The unique geospatial capabilities of Instagram are enhanced by the fact that users often post at sites such as restaurants, bars, workplaces, or small businesses within the known radius of transit system stops. Although these users are not posting about public buses or trains directly, their content helps more thoroughly map the commercial and cultural geographies of interest to transit stakeholders. Small businesses, moreover, are eager to engage transit riders, and a smart social media strategy may pay valuable dividends in terms of partnerships and sponsorships with local vendors.
Besides their photographic image content, Instagram posts generate text useful for machine learning applications. In the course of our pilot project, a visitor to the city recorded her “pleasant surprise” at “how easy it was to catch light rail” from the airport. With textured language input such as this, sophisticated data analysis programs can assign a sentiment score to the post. By gauging favorable sentiments across a wide range of data, communication departments within your company can view trends over time and better respond to their customer base’s overall perceptions.
To date, 40 billion photos have been shared on Instagram. Advances in machine learning promise to open new frontiers for practical analysis and deployment of such content. Such finely-resolved user information might have been the dream of a previous generation of transit planners; the sheer volume and particularity of data on demographics, logistics, user behavior, and subjective reporting holds the potential to revolutionize how we think of the enterprise of mass transit. It falls to a new cohort of stakeholders to utilize the full potential of such technology.
Scott Wilkinson is president of AlphaVu.
