Intelligence Cycle
The conceptual model of how intelligence is produced — planning, collection, processing, analysis, dissemination
Audio readout of this entry.
The intelligence cycle is the conceptual model used in intelligence-studies literature and in service training to describe how intelligence is produced. The conventional sequence comprises five or six steps: planning and direction (the establishment of intelligence requirements by policy customers), collection (the gathering of information through HUMINT, SIGINT, IMINT, MASINT, OSINT, or GEOINT), processing and exploitation (translation, decryption, photographic interpretation, technical analysis), analysis and production (the synthesis of collected material into finished intelligence products), dissemination (distribution to consumers), and feedback (consumer response that shapes the next cycle).
The cycle is useful as an introductory frame because it captures the basic logic of the enterprise: requirements drive collection, collection produces raw material, analysis produces finished product, finished product informs decisions, decisions generate new requirements. It also makes visible the parts of the workflow that are not collection — processing, analysis, dissemination — that account for a large fraction of an intelligence service's institutional capacity and that determine whether collected material becomes useable intelligence at all.
The actual workflow is messier than the cyclical diagram suggests. Requirements are continuously revised rather than set at the start of a cycle. Collection is opportunistic: a service rarely turns off a productive channel because its formal requirement has expired. Analysts work from streams that arrive on independent schedules; dissemination is shaped as much by consumer politics as by formal taskings. The intelligence-failure literature — on Pearl Harbor, on the 1973 Yom Kippur attack, on 9/11, on Iraqi WMD, on the October 2023 Hamas attack — repeatedly identifies failures of analytical synthesis and consumer reception rather than failures of collection. The cycle diagram does not, by itself, explain why those failures happen.
The dossiers on this site that document intelligence failures — the October 7 Warning Failure most directly, with broader treatment in the methodology page's discussion of selection bias in the declassified record — illustrate the gap between the schematic cycle and the institutional reality. The cycle is the right place to start; it is not the place to stop.