Votta, F., Dobber, T., Guinaudeau, B., Helberger, N., & de Vreese, C. (2025). The cost of reach: Testing the role of ad delivery algorithms in online political campaigns. Political Communication, 42, 476–508. https://doi.org/10.1080/10584609.2024.2439317
Summary
This paper examines how ad delivery algorithms — the platform-side systems that decide who actually sees a political ad and at what price — shape the reach and cost of online political campaigns on platforms such as Meta and Google. Moving beyond the familiar critique of advertiser-side microtargeting, the authors argue that platforms themselves make consequential, largely opaque decisions about audience delivery that campaigners cannot fully control and regulators cannot easily observe. In doing so, the paper reframes debates over political advertising toward the delivery stage of the pipeline, situating algorithmic mediation as a distinct locus of platform power in democratic discourse.
Key Contributions
- Provides empirical evidence that ad delivery algorithms, not just advertiser targeting choices, materially influence which audiences see political ads and what they cost.
- Highlights the opacity of delivery decisions to campaigners, voters, and regulators alike.
- Bridges political communication scholarship with platform governance and algorithmic accountability literatures.
- Informs ongoing regulatory and transparency debates about political ad delivery systems.
Methods
- Empirical analysis of online political advertising data, drawing on platform ad libraries (e.g., Meta, Google).
- Tests how delivery algorithms affect ad cost and audience reach metrics, isolating the “cost of reach” as an algorithmically mediated outcome rather than a purely advertiser-determined one.
Findings
- Ad delivery algorithms substantively shape campaign reach and cost dynamics beyond advertiser targeting.
- These delivery effects operate opaquely, leaving campaigners, audiences, and regulators without clear visibility into how outcomes are produced.
- (Detailed quantitative results are not recoverable from the abstract alone.)
Connections
This paper shares an author with de-Vreese2026-zx and sits within the platform-governance and elections literatures, speaking directly to work on transparency and access to platform ad data such as Rieder2025-ju and Rieder2026-pp. Its focus on how platform algorithms mediate political reach connects to broader studies of algorithmic influence and platform power in electoral contexts, including Gonzalez-Bailon2024-rq and Bechmann2026-dr.