{"site":{"name":"Koji","description":"AI-native customer research platform that helps teams conduct, analyze, and synthesize customer interviews at scale.","url":"https://www.koji.so","contentTypes":["blog","documentation"],"lastUpdated":"2026-05-24T11:39:08.186Z"},"content":[{"type":"documentation","id":"55f6049b-4212-46ad-92b9-b0b10ddd84de","slug":"aeiou-framework-ux-observation","title":"The AEIOU Framework: How to Structure Field Observations for UX Research (2026 Guide)","url":"https://www.koji.so/docs/aeiou-framework-ux-observation","summary":"AEIOU is a five-letter observation framework — Activities, Environments, Interactions, Objects, Users — developed in 1991 at Doblin by Rick Robinson, Ilya Prokopoff, John Cain, and Julie Pokorny. It gives field researchers a mutually exclusive, collectively exhaustive (MECE) taxonomy for capturing and coding observations during ethnographic studies and contextual inquiry. Activities are goal-directed actions; Environments are the spaces; Interactions are person-to-person or person-to-system exchanges; Objects are tools and materials; Users are the people themselves. The 5-step workflow: define the research question, recruit and consent, observe and code live, debrief within 24 hours, synthesize across participants. Modern adaptation pairs AEIOU with AI-moderated interviews (Koji) to extend the framework to remote and hybrid research, with adaptive branching capturing the micro-activity reasoning a passive observer would miss.","content":"# The AEIOU Framework: How to Structure Field Observations for UX Research (2026 Guide)\n\n**Bottom line:** AEIOU is a five-letter observation framework — Activities, Environments, Interactions, Objects, Users — that gives researchers a mutually exclusive, collectively exhaustive (MECE) structure for capturing field data. Developed at Doblin in 1991 by Rick Robinson, Ilya Prokopoff, John Cain, and Julie Pokorny, AEIOU remains one of the most widely taught scaffolds in ethnographic and contextual UX research. Used well, it turns the chaos of a field study into coded data ready for synthesis. Used poorly, it becomes a clipboard exercise that misses the point of being in the field.\n\nThis guide is for UX researchers, design strategists, and PMs who run field studies, contextual inquiries, or any observational research. We cover the origin of the framework, what each letter actually means in practice, when to use AEIOU versus alternatives, the field-ready note-taking template, the synthesis workflow, and how AI-moderated interviews can extend the framework to remote and distributed research.\n\n## Where AEIOU Came From\n\nAEIOU was created in 1991 at Doblin, a strategy and design consultancy, by Rick Robinson, Ilya Prokopoff, John Cain, and Julie Pokorny. Their goal was practical: ethnographers and design researchers were drowning in unstructured field notes, and synthesis sessions kept collapsing into \"everyone remembers a different study.\" The team needed a taxonomy that was mutually exclusive (each observation goes in exactly one bucket) and collectively exhaustive (any meaningful observation has a bucket).\n\nThe framework moved with Rick Robinson to E-Lab in the late 1990s and was published in the EPIC community's design research literature, where it became the de facto vocabulary for entry-level field research. The IIT Institute of Design later adopted AEIOU as a teaching scaffold, which is why some sources misattribute its origin there.\n\nThe five letters have survived three decades of research-methods evolution because they map cleanly to how people actually move through the world: people *do things* (activities), *somewhere* (environments), *with someone* (interactions), *using something* (objects), and the *people themselves* matter (users).\n\n## What Each Letter Captures\n\n### A — Activities\n\nActivities are goal-directed actions and the paths people take to accomplish them. They are bigger than micro-gestures but smaller than life narratives. Watch for the sequence of steps, the modes of work (focused, distracted, collaborative, interrupted), and the rituals — the things people do every time without conscious thought.\n\n*Example prompt:* What is the user doing right now to make progress toward their goal? What sequence of steps does the activity follow?\n\n### E — Environments\n\nEnvironments are the physical, digital, or social spaces in which the activities take place. Capture the layout, lighting, noise, temperature, and the implicit \"rules\" of the space. Environments often constrain what activities are possible — a kitchen during dinner prep affords different behaviors than the same kitchen at 7am.\n\n*Example prompt:* Where is this happening? What does the space afford or prevent? Who else uses this space?\n\n### I — Interactions\n\nInteractions are the exchanges between people and other people, or between people and systems. They are the building blocks of activities. Distinguish synchronous (live conversation, shoulder-tapping) from asynchronous (email, comments, slacks); intentional from incidental.\n\n*Example prompt:* Who or what is the user talking to or coordinating with? How does information flow between them?\n\n### O — Objects\n\nObjects are the building blocks of environments — tools, devices, materials, documents, displays. Watch how objects are used in their intended way and, more importantly, how they are repurposed. The post-it stuck to a monitor, the manual annotated in pencil, the second phone used only for one app — these workarounds are gold.\n\n*Example prompt:* What tools or materials are present? Are they used as designed, or repurposed?\n\n### U — Users\n\nUsers are the people whose behavior, preferences, motivations, and circumstances anchor the entire study. Capture their role, the relationships they have to others in the scene, their stated goals, and the emotions visible in their body language and word choice.\n\n*Example prompt:* Who is this person in this context? What do they care about? What is their relationship to others in the scene?\n\n## When to Use AEIOU (And When Not To)\n\n**Use AEIOU when:**\n\n- You are running a [contextual inquiry](/docs/contextual-inquiry) or [field study](/docs/ethnographic-research) and need a structured note-taking scaffold\n- Multiple researchers are running parallel sessions and you need a shared vocabulary so notes can be merged\n- You are training new researchers and want a framework that prevents the \"I just sat there and watched\" trap\n- You are doing [observational research](/docs/observational-research-guide) to inform a new product category where you cannot rely on existing user mental models\n\n**Use a different framework when:**\n\n- You are running task-based usability testing — use [the 5-second test](/docs/5-second-test-guide) or [tree testing](/docs/tree-testing-guide) instead\n- You need to capture longitudinal behavior change — use [diary studies](/docs/diary-study-guide)\n- Your research question is about attitudes only, not behavior — see [attitudinal vs. behavioral research](/docs/attitudinal-vs-behavioral-research)\n- The setting is fully remote and you cannot observe environments or objects directly — use AEIOU adapted (covered below) or shift to a journey-mapping protocol\n\n> \"Field studies to observe users in their natural habitat are one of the most important user research methods… through observation and collaborative interpretation, contextual inquiry uncovers hidden insights about customers' work that may not be available through other research methods.\"  \n> — **Nielsen Norman Group**\n\nThe strength of AEIOU is exactly here — it gives field researchers the structure that prevents observation from collapsing into impressionistic memory.\n\n## The Field-Ready AEIOU Template\n\nCarry this template into every session, paper or digital. Each row is one observation. Coding happens in the field, not after.\n\n| Time | A — Activity | E — Environment | I — Interaction | O — Object | U — User |\n|------|--------------|-----------------|------------------|------------|----------|\n| 09:14 | Annotating a printed customer journey map | Open-plan kitchen-table workspace | Calls partner over to point at sticky note | Printed map, red pen, three sticky-note colors | Senior PM, 8 yrs experience, visibly tired |\n| 09:21 | Switches to laptop to look up customer name | Same kitchen table | Slacks designer for context | MacBook, Slack open, Notion second tab | Same PM |\n\nThe temporal column matters more than people expect. The rhythm of activity transitions is itself a finding — frequent switching reveals interruption load; long uninterrupted blocks reveal flow.\n\nA common variant called **AEIOU+E** adds **Emotions** as a sixth letter, tagging affect alongside behavior. Useful for service design and any context where the emotional arc is the deliverable.\n\n## The 5-Step AEIOU Field Study Workflow\n\n### Step 1: Define the Research Question\n\nAEIOU is a structuring tool, not a question generator. Before any field visit, write a one-sentence research question. *Example:* \"How do small-team PMs decide which customer feedback to act on each week?\"\n\nIf you cannot state the question, your field notes will sprawl. The question scopes which activities, interactions, and objects matter.\n\n### Step 2: Recruit and Schedule\n\nRecruit 5-8 participants whose context matches your research question. Use [screener questions](/docs/research-screener-questions) to confirm fit, and follow the [research consent](/docs/research-consent-form-templates) protocol — field studies involve recording in someone's space, which raises consent and privacy stakes beyond a standard interview.\n\n### Step 3: Observe and Code Live\n\nSpend 60-120 minutes with each participant in their actual context — desk, kitchen, workshop, store, vehicle. Two roles work best: one researcher engages with the participant; one researcher captures AEIOU rows silently. After 20 minutes, the engaging researcher pauses for \"think-aloud\" prompts: \"Walk me through what you just did. Why that sticky note and not a Notion comment?\"\n\nThe non-negotiable rule: code in the field, in the moment. Notes coded after the fact lose the precision that makes AEIOU worth using.\n\n### Step 4: Debrief Within 24 Hours\n\nSchedule a 60-minute team debrief within one day of the session while observations are fresh. Walk every row of the AEIOU table. Mark surprises — the things you expected and did not see, the things you saw and did not expect. Surprises are where insight lives.\n\n### Step 5: Synthesize Across Participants\n\nAfter all sessions, lay every AEIOU row across all participants on a shared board. Cluster by column — all activities together, all environments together, etc. Affinity-map within each column. Cross-tag patterns that span columns (e.g., a recurring object that appears in three different environments).\n\nThis is the step that transforms field notes into a [thematic analysis](/docs/thematic-analysis-guide) that product teams can act on.\n\n## The Limitations of AEIOU (Worth Knowing Before You Adopt It)\n\nAEIOU is a behavioral observation tool, not a complete research method. Three known limits:\n\n1. **It captures what is observable, not what is felt.** Goals, frustrations, and unmet needs require complementary interviews — pair AEIOU with [empathy interviews](/docs/empathy-interview-guide) or [jobs-to-be-done interviews](/docs/jobs-to-be-done-interviews) to triangulate.\n2. **It assumes co-presence.** Original AEIOU was designed for in-person field studies. Remote-first research breaks \"Environments\" and \"Objects\" because you can only see what the camera sees.\n3. **It can become a checklist.** Junior researchers sometimes fill the table mechanically without asking *why* the patterns matter. Add a sixth column — \"So what?\" — to force interpretation in the moment.\n\n## Modern Approach: AEIOU at Remote Scale With AI Interviews\n\nThe original 1991 method assumed you and the participant were in the same room. In 2026, most product research is hybrid or fully remote — which appears to break AEIOU. It does not. The framework adapts cleanly.\n\nHere is how Koji extends AEIOU to remote and distributed research:\n\n**A — Activities** become tasks the participant narrates and demonstrates over screen-share. Koji's [adaptive AI interview branching](/docs/adaptive-ai-interview-branching) follows up live: \"You just paused before clicking Send — what made you hesitate?\" — capturing the micro-activity reasoning that a static script would miss.\n\n**E — Environments** become the participant's described environment (\"I usually do this between standups, on my phone, on the train\"). The AI interviewer probes for environmental context using [structured questions](/docs/structured-questions-guide) — including `single_choice` and `multiple_choice` types that pre-code environmental segments for synthesis.\n\n**I — Interactions** are captured by asking about coordination patterns (\"Who do you involve when you make this decision?\"). Koji surfaces these as relationship maps during [thematic analysis](/docs/thematic-analysis-guide).\n\n**O — Objects** become the tools, documents, and apps the participant references. AI follow-up makes object identification cleaner than a passive observer ever could: \"What is the Notion template you mentioned — is that yours or something your team uses?\"\n\n**U — Users** are richly profiled by Koji's screening logic and structured questions, automatically creating segment cohorts for downstream synthesis.\n\n### Why This Matters\n\nThe traditional bottleneck in AEIOU-based research is synthesis. A team running 8 field studies generates 8 hours of recordings, several hundred AEIOU rows, and a multi-week affinity-mapping marathon. Teams using AI-assisted research tools report 60% faster time-to-insight — for AEIOU studies specifically, that means going from raw notes to themed insights in days instead of weeks.\n\nWhen the synthesis bottleneck breaks, two things change. First, sample sizes can grow — instead of 8 deep field studies you can run 30 hybrid sessions. Second, the analysis becomes continuous rather than episodic, fitting into [continuous discovery](/docs/continuous-discovery-tools-2026) rhythms instead of being a once-per-quarter event.\n\n## AEIOU vs. Other Field-Research Frameworks\n\n| Framework | Best For | Limitation |\n|-----------|----------|------------|\n| **AEIOU** | Structured observation across multiple sessions and researchers | Behavioral, not attitudinal |\n| **POEMS** (People, Objects, Environments, Messages, Services) | Service-design field studies | More complex; longer to teach |\n| **9 Dimensions of Observation** (Spradley) | Academic ethnography | Heavier framework, less practical for sprints |\n| **Empathy Maps** | Synthesizing observations into design-ready personas | Output format, not a field-capture method |\n| **Customer Journey Mapping** | Showing the temporal arc of an experience | Better for synthesis than for raw capture |\n\nPair AEIOU with one of these for the full pipeline: AEIOU captures, [empathy maps](/docs/empathy-map-guide) synthesize, [customer journey maps](/docs/customer-journey-mapping) communicate.\n\n## Common AEIOU Mistakes\n\n**Mistake 1: Coding after the session.** Memory degrades within minutes. Codes filled in two hours later are reconstructions, not observations.\n\n**Mistake 2: Forcing every observation into one bucket.** Some observations belong in two columns (e.g., a coffee cup is an Object and a hint about the Environment's affordances). Allow cross-tagging.\n\n**Mistake 3: Ignoring the \"+Emotions\" extension.** For any consumer or service-design study, the affect dimension is too important to drop.\n\n**Mistake 4: Treating AEIOU as the whole methodology.** It is a capture scaffold. It does not replace interview design, sampling rigor, or [research planning](/docs/ux-research-plan-template).\n\n**Mistake 5: Using AEIOU when a [usability test](/docs/usability-testing-guide) is the right method.** AEIOU is for understanding behavior in context — not for evaluating a specific interface against task success criteria.\n\n## Key Statistics and Sources\n\n- AEIOU was developed in 1991 by Rick Robinson, Ilya Prokopoff, John Cain, and Julie Pokorny at Doblin ([EPIC People](https://www.epicpeople.org/building-a-useful-research-tool/))\n- AEIOU is designed to be MECE — mutually exclusive, collectively exhaustive — across its five categories, which is what enables multi-researcher synthesis\n- Nielsen Norman Group classifies field studies as \"one of the most important user research methods\" specifically because of their power to surface hidden work practices ([NN/g — Field Studies](https://www.nngroup.com/articles/field-studies/))\n- Teams using AI-assisted research tools report **60%** faster time-to-insight versus traditional moderated synthesis — a decisive advantage for AEIOU studies, where synthesis is historically the bottleneck\n\n## Related Resources\n\n- [Structured Questions Guide: 6 Question Types for Sharper Research](/docs/structured-questions-guide)\n- [Contextual Inquiry: The Complete Guide to Observational Research](/docs/contextual-inquiry)\n- [Ethnographic Research: Methods, Examples, and UX Applications](/docs/ethnographic-research)\n- [Observational Research: How to Learn From What Users Do, Not What They Say](/docs/observational-research-guide)\n- [Behavioral Research Methods: The Complete Guide for Product and UX Teams](/docs/behavioral-research-methods)\n- [Empathy Map: The Complete Guide to Building User Empathy](/docs/empathy-map-guide)\n- [Customer Journey Mapping: The Complete Guide for UX Teams](/docs/customer-journey-mapping)\n- [Attitudinal vs. Behavioral Research: What Users Say vs. What They Do](/docs/attitudinal-vs-behavioral-research)\n- [Diary Studies: The Complete Guide to Longitudinal User Research](/docs/diary-study-guide)\n- [Thematic Analysis: The Complete Guide](/docs/thematic-analysis-guide)\n","category":"Research Methods","lastModified":"2026-05-24T03:36:48.314917+00:00","metaTitle":"AEIOU Framework: The Complete Guide to Field Observation in UX Research (2026) | Koji","metaDescription":"The AEIOU framework structures field observations across Activities, Environments, Interactions, Objects, and Users. Learn the origin, field-ready template, 5-step workflow, comparisons to POEMS and empathy maps, and how AI-moderated interviews extend AEIOU to remote research.","keywords":["aeiou framework","aeiou design thinking","aeiou ux research","aeiou observation framework","field observation framework","contextual inquiry framework","doblin aeiou","ethnographic research framework","observation method ux","activities environments interactions objects users","aeiou template","aeiou example"],"aiSummary":"AEIOU is a five-letter observation framework — Activities, Environments, Interactions, Objects, Users — developed in 1991 at Doblin by Rick Robinson, Ilya Prokopoff, John Cain, and Julie Pokorny. It gives field researchers a mutually exclusive, collectively exhaustive (MECE) taxonomy for capturing and coding observations during ethnographic studies and contextual inquiry. Activities are goal-directed actions; Environments are the spaces; Interactions are person-to-person or person-to-system exchanges; Objects are tools and materials; Users are the people themselves. The 5-step workflow: define the research question, recruit and consent, observe and code live, debrief within 24 hours, synthesize across participants. Modern adaptation pairs AEIOU with AI-moderated interviews (Koji) to extend the framework to remote and hybrid research, with adaptive branching capturing the micro-activity reasoning a passive observer would miss.","aiPrerequisites":["Basic familiarity with qualitative UX research","Understanding of what a contextual inquiry or field study is"],"aiLearningOutcomes":["Identify the five AEIOU categories and what each captures in practice","Decide when AEIOU is the right framework versus alternatives like POEMS or empathy mapping","Run a 5-step AEIOU field study workflow from research question to synthesis","Recognize the five most common AEIOU mistakes that flatten insight quality","Extend AEIOU to remote and hybrid research using AI-moderated interviews"],"aiDifficulty":"intermediate","aiEstimatedTime":"17 minutes"}],"pagination":{"total":1,"returned":1,"offset":0}}