ensure the quality of third-party apps, but also came at a time when Instagram itself was beginning to undergo significant change and internal consolidation (Constine 2015). While the API changes did not have the same impact as they did when Twitter made the first dramatic changes to their API (Bohn 2012), it nevertheless demonstrated a significant shift in the way Instagram interacted with other developers who utilized the Instagram platform in any capacity. Instagram’s API changes essentially ended an early phase where third-party applications had almost free reign, to a more controlled ecology where Instagram were far more careful and selective about who could access Instagram data and what could be done with it.
It is no coincidence that Instagram’s API changes came in the midst of the platform releasing several standalone apps of its own. In August 2014, Instagram quietly released the Hyperlapse app for Apple devices, which allowed users to speed up, smooth and stabilize video footage to create appealing, condensed videos of longer periods of activity. In March 2015, another Instagram app, Layout, was added, which allowed users to create image collages of various forms and post these directly to Instagram. Layout was particularly notable in that it replicated the functionality present in many of the third-party Instagram apps, showing that Facebook was all too happy to pick and choose some of the most popular elements of Instagram’s ecology of third-party apps. The third Instagram-related app, Boomerang, was released in October the same year; it allowed the capture of two-second long looped videos, in aesthetically similar territory to animated GIFs (Miltner & Highfield 2017), but as video rather than image files. The emergence of Instagram’s own suite of apps, and the changes forced onto all third-party apps accessing Instagram’s data, is explored in more detail in chapter 3. Of course, the most turbulent relationship between Instagram and another platform is the ongoing turf war with Snapchat, which is explored below in relation to the emergence of Stories.
Finstagrams, Rinstagrams and Multiple Accounts
While Mark Zuckerberg has been widely and publicly insistent that Facebook users only have one ‘real’ identity, and should thus only have one ‘real’ Facebook account (Haimson & Hoffmann 2016), Instagram has historically been far more flexible in terms of names, identities and multiple accounts. However, while uniqueness might not have characterized all Instagram users, relatively quickly a certain aesthetic expectation did: Instagram feeds were highly polished and curated, not really spaces for untamed frivolity or silliness. In this context, Finstagrams, or Finstas, emerged; Finstagrams were almost always private accounts, with very low numbers of (trusted) followers, featuring content that was often disruptive, ironic or at odds with the main Instagram aesthetic (such as reposted content, memes and so forth). These led to the main Instagram accounts being redubbed Rinstagrams, or Rinstas, which, were usually public accounts, featuring polished content that were carefully edited to add to a specific look, style or brand. Yet Finstas, to some extent, were swiftly seen as more authentic: in a New York Times piece (Safronova 2015) reporting on the popularity of Finstas, these ‘Fake Instagram’ accounts were specifically characterized as presenting something portrayed as lacking in ‘real’ Instagram: ‘truer version[s] of themselves than their main profiles’. Seemingly endorsing the holding of multiple accounts, addressing different audiences, in February 2016 Instagram added the ability to move between multiple Instagram accounts without having to log out. Switching between Instas and Finstas became a breeze. While Instagram’s owner Facebook consistently argues that there is only one ‘real’ authentic ‘you’ online (van Dijck 2013), Instagram’s promotion of the ease of movement between multiple accounts, and the emergence of prompts to ‘Share a Different Side of Yourself. Create a private account to share photos and videos with a close group of followers’ (see figure 1.1), reveals Instagram is far more supportive of the idea that multiple accounts are needed to address and engage with different groups and contexts (Abidin 2017d).
Figure 1.1. Instagram ‘Create New Account’ popup message, 2017
While one motivation for Instagram’s flexibility is clearly financial – multiple accounts almost certainly means more time on Instagram, more time seeing advertisements and more loyalty to the platform – it is also important to note that the platform is simultaneously more flexible and accommodating to a multiplicity of voices, perspectives and performances by individuals and groups. It’s also a big bonus for social media managers, of course, meaning that multiple people can contribute to and manage a particular brand’s account without having to completely log out of their own personal accounts every time. While Facebook’s real identity policy faced a considerable backlash from queer communities (Newton 2014) and other marginalized groups for whom digital pseudonymity is paramount for personal safety and self-actualization (Holpuch 2015), the company responded to criticism minimally, insisting identities were still singular, but relaxing their policy to allow users to use ‘the name they go by in everyday life’ to ‘keep our community safe’ (Facebook 2018b). On Instagram, multiple accounts are not just tolerated, but actively encouraged, which allows for far more meaningful privacy and audience segmentation. Of course, where privacy is a concern, Finstas are an important strategy for controlling who sees what content on Instagram (Duffy & Chan 2019), yet as early as 2015 popular reporting noted the problem of ‘finsta snitches’ (Safronova 2015) who betrayed the trust of Finsta users, capturing and reposting Finsta content in more mainstream settings, disrupting the users’ attempts to control the contexts in which their content is seen. While Finstas and multiple accounts are thus not perfect in terms of allowing private and more carefully moderated spaces, Instagram’s approach nevertheless is far more flexible and meaningful than is possible on Facebook’s main platform (Highfield & Leaver 2015).
The Algorithmic Timeline
In June 2016, Instagram shifted from a chronological display of Instagram content in a user’s feed to an algorithmically sorted timeline in a move which angered many users and was widely considered more like Mark Zuckerberg’s famous instruction to ‘move fast and break things’ (Vaidhyanathan 2018). Like the Facebook newsfeed and related proprietary tools for making decisions about content and curation, the algorithmic timeline is opaque to most users, which led to wild speculation about exactly how decisions are made. In 2018 Instagram briefed technology reporters on the broad parameters of how the algorithmic timeline operates, in part to reassure businesses and brands that the algorithms would continue to ensure a level playing field and that users would see their content, even as the number of users continues to grow. According to their briefing, each user’s unique Instagram feed is based on three core categories:
Interest – how much Instagram perceives a user will want to see a post based on past viewing of similar content;
Recency – how new the post is; and
Relationship – how close a user is to the user posting the content. This is determined by a range of things, including frequency of past liking, comments and being tagged in photos together.
In addition, other minor signals that influence the timeline include how often a user opens Instagram, how many people they follow, and how much time they tend to spend on Instagram each time it is opened (Constine 2018b).
Of course, there are many other algorithms at work on Instagram, from those which determine suggested accounts to follow, through to those that flag content for moderation or removal, through to those that curate the Explore area, matching content and accounts with the recorded activity of each user. Indeed, even the previous chronological timeline was delivered by an algorithm, although the difference here is that the operation of that algorithm was transparent to users. While the scope of algorithmic activities are often invisible to users, and
