HKSM

A data-gathering technique that combines watching real work as it happens with direct conversation to understand what is occurring and why. It helps reveal tacit knowledge, workarounds, and context that documents and metrics may miss.

Key Points

• Blends direct observation with informal or structured conversation to validate what is actually happening.
• Effective for uncovering tacit knowledge, workarounds, and differences between documented and real processes.
• Plan the approach to minimize disruption and bias, and obtain consent when needed.
• Useful across discovery, requirements, process improvement, risk identification, and verification of changes.
• Notes should be factual and time-stamped, with findings later validated by participants.
• Works best when triangulated with other techniques such as interviews, document analysis, and metrics.

When to Use

• Early discovery or requirements elicitation in unfamiliar or complex workflows.
• To verify actual process performance, handoffs, bottlenecks, and non-value-add activities.
• When reported metrics conflict with stakeholder claims or documentation.
• For user experience studies, safety-critical tasks, or high-variability work.
• During change impact analysis, pilot trials, or post-implementation adoption checks.

How to Use

1. Define purpose, scope, questions, and success criteria for the observation.
2. Identify participants and locations; obtain permissions and address privacy and confidentiality.
3. Select an approach: shadowing, contextual inquiry, time-and-motion sampling, gemba walk, or remote observation.
4. Prepare tools: observation guide or checklist, conversation prompts, consent script, and recording templates.
5. Conduct observation with minimal interference; capture objective notes, timestamps, metrics, and artifacts.
6. Ask open-ended questions at natural pauses to clarify intent, constraints, and pain points.
7. Synthesize notes into themes, process maps, and candidate requirements, risks, and issues.
8. Validate findings with participants and sponsors; finalize documented outputs and next steps.

Inputs Needed

• Objectives, scope, and an observation guide or research plan.
• Stakeholder register, roles, access permissions, and schedule.
• Existing process maps, SOPs, system screenshots, and performance metrics.
• Privacy, legal, and ethics guidance, including NDAs or consent requirements.
• Recording tools such as note templates, time trackers, and approved audio/video tools.

Outputs Produced

• Observation notes, conversation summaries, and validated findings.
• Documented requirements, user stories, use cases, and acceptance criteria.
• Updated process maps or value stream maps with actual steps, waits, and defect points.
• Identified risks, issues, impediments, and improvement opportunities.
• Change requests, backlog items, lessons learned entries, and stakeholder updates.

Example

A project team is improving a service request workflow. The PM and business analyst shadow agents for several shifts, quietly timing key steps and noting handoffs. At natural breaks, they ask brief questions to understand why agents use certain shortcuts. The team consolidates notes into a current-state map, identifies two redundant approvals and a tool login delay, validates findings with agents and a supervisor, and creates backlog items to remove one approval and enable single sign-on.

Pitfalls

• Hawthorne effect: people change behavior when observed; mitigate by extending observation and being unobtrusive.
• Confirmation bias: seeing only what you expect; mitigate with a structured guide and peer review of notes.
• Interrupting the work and influencing outcomes; ask questions at natural pauses and avoid coaching during observation.
• Insufficient sample size or narrow time window; include variations such as shifts, peak periods, and different roles.
• Privacy or compliance breaches; follow policy, obtain consent, and avoid capturing sensitive data.
• Failing to validate interpretations; review summaries with participants before finalizing outputs.

Related Items

• Interviews and one-on-one conversations.
• Contextual inquiry and job shadowing.
• Gemba walk and process walk-throughs.
• Workshops and facilitated sessions.
• Document analysis and value stream mapping.
• Surveys/questionnaires and usability testing.
• Prototypes and simulations.

PMP Example Question

A PM suspects the documented process does not match actual practice in a high-volume operations area. What is the best next step to understand the real workflow and its pain points?

1. Send a survey asking workers to rate each step and propose improvements.
2. Update the SOP based on management input and roll out training.
3. Perform observation/conversation by shadowing staff and asking clarifying questions at natural breaks.
4. Schedule a lessons learned meeting with the team leads only.

Correct Answer: C — Perform observation/conversation by shadowing staff and asking clarifying questions at natural breaks.

Explanation: Direct observation with conversation reveals actual behaviors and tacit knowledge, reducing bias and validating how work is truly performed. Surveys or SOP updates alone may miss context and workarounds.