Bucher, T. (2017). The algorithmic imaginary: exploring the ordinary affects of Facebook algorithms. Information, Communication & Society, 20, 30–44. https://doi.org/10.1080/1369118x.2016.1154086
Summary
Bucher investigates how ordinary Facebook users come to notice, feel, and make sense of algorithms that are by design hidden and invisible. Drawing on tweets and email interviews with 25 users, she develops the concept of the algorithmic imaginary — the ways people imagine what algorithms are, what they should be, and how they function. Rather than treating these imaginings as mistaken folk beliefs, she argues they are affectively charged and generative: they shape user behavior, which in turn feeds back into and reshapes the algorithm itself. The core move is to shift critical algorithm studies away from opening the technical black box and toward understanding how algorithms make people feel, treating affect as central to algorithmic social power.
Key Contributions
- Introduces and theorizes the algorithmic imaginary as a framework for lived, affective encounters with algorithms.
- Shifts the agenda from algorithm awareness (whether users know algorithms exist) to how users experience and interpret algorithmic encounters in daily life.
- Offers a method for accessing personal algorithm experiences via social media traces plus interviews — “getting behind the tweets” rather than reverse-engineering code.
- Advances an affect-theoretical and phenomenological approach to critical algorithm studies.
- Reframes algorithmic power as recursive and co-constituted between people and systems.
Methods
Qualitative study combining analysis of public tweets with email interviews of 25 ordinary Facebook users (from 47 contacted). Over nine months (Oct 2014–June 2015), Bucher ran regular Twitter searches on keyword combinations such as “Facebook algorithm,” “algorithm AND weird,” and “algorithm AND creepy,” manually scrolling and screenshotting personal (non-marketing) tweets. Short email interviews then probed context, perceived workings, and effects on platform use. Responses were coded for triggering situations, mental models, effects, tactics, and concerns. The theoretical framing draws on affect theory (Gregg & Seigworth; Deleuze & Guattari), Berlant’s “the situation,” Stewart’s Ordinary Affects, and Merleau-Ponty’s phenomenology.
Findings
- Algorithms become noticeable mainly in moments of breakdown, surprise, or violated expectation.
- Profiling can feel uncomfortable even when accurate (e.g., being pegged as “pregnant, single, broke”; age-stereotyped ads).
- Uncanny “whoa moments” arise when ads reflect immediate offline behavior (an air mattress ad, Nespresso ads).
- “Faulty prediction” breakdowns provoke frustration when the feed fails to reflect a user’s current identity (“out of sync”).
- Many users treat Facebook as a “popularity game,” developing optimization tactics — timing, word choice, multiple photos — to be recognized.
- Features produce “cruel connections,” e.g., a “Year in Review” surfacing a deceased daughter, foregrounding perceived algorithmic inhumanity.
- Users feel the algorithm reshapes friendships, hiding and resurfacing people and eroding a sense of control.
- Strategic reactions span optimization, obfuscation, and deliberately “confusing” the algorithm — which then feeds back into the system.
Connections
This paper is foundational for work on how users perceive, theorize, and strategically respond to opaque ranking systems; it connects directly to studies of algorithmic “folk theories” and platform literacy in Gagrcin2024-dl and Cotter2019-js, and to the professional optimization logics documented in Petre2019-zh. Its argument that algorithms are co-constituted through user behavior complements platform-governance and visibility accounts such as Gillespie2022-jx, while its concern with felt but accurate profiling and ad experiences resonates with the engagement-and-exposure questions examined in Bakshy2015-rn.