Algorithmic opacity and the distortion of B2B signalling: evidence from digital marketing client-agency services

Purpose The paper aims to examine how algorithmic opacity in digital platforms reshapes signalling processes within client-agency relationships. While signalling theory assumes that agencies can convey qualities such as expertise, performance outcomes and accountability through costly, observable and reliable signals, digital marketing services are increasingly enacted within platform-mediated environments that distort what information can be produced, accessed and evaluated by the client. Design/methodology/approach Data were collected through 32 interviews with clients, agency representatives, digital marketing bodies, procurers, trainers and legal professionals. Interviews were thematically analysed in NVivo using Braun and Clarke’s (2006) method. Findings The study identifies three interrelated signalling distortions: Influence, Gatekeeper and Proof, that explain how platform dynamics alter the conditions under which signals are produced, observed and interpreted. Platforms limit pertinent algorithm update information, restrict access to performance data and complicate attribution of outcomes, thereby destabilising the core signalling conditions of costliness, observability and reliability. Signals traditionally used to convey competence and accountability, such as performance reports or case evidence, lose their differentiating power when evaluation is mediated by opaque algorithmic systems. Originality/value The paper addresses the call of scholars who argue for the use of signalling theory in contemporary contexts, to show how the client-agency signalling environment is now platform-mediated and distorts the informational conditions of signalling. A new algorithmic risk disclosure signal is introduced, outlining a proactive mechanism for agencies to disclose algorithmic risks to clients.