Observation Methods: Seeing Your Process in Action

Direct observation reveals what documentation misses and interviews can't capture. This chapter covers practical techniques for watching processes unfold in real-time and extracting meaningful insights.

Why Observation Matters

People often describe what should happen rather than what actually happens. Observation closes this gap by revealing:

• Actual behaviors vs. documented procedures
• Workarounds people have developed
• Environmental factors that affect performance
• Informal communication and coordination
• Physical constraints that limit options

"Observations serve two purposes: discovering procedures qualitatively and determining operational parameters quantitatively."

Types of Walk-throughs

Guided Walk-throughs

A subject matter expert leads you through the process, explaining each step.

When to use:

• You're new to the process
• Complex technical operations
• Safety considerations
• Need to understand why things happen

Best practices:

• Prepare questions in advance
• Ask "what happens if...?" to explore exceptions
• Request demonstrations, not just explanations
• Note where the expert hesitates or qualifies

Example:

"During the insurance underwriting tour, the senior adjuster walked me through a claim from intake to resolution. She explained the decision criteria at each step and showed me the screens she uses. When we got to the fraud detection step, she mentioned that experienced adjusters often notice patterns the system misses—this informal knowledge wasn't in any procedure."

Unguided Walk-throughs

You observe independently, following the work as it flows.

When to use:

• You want to see unfiltered behavior
• Checking if guided tours match reality
• Understanding flow and timing
• Identifying inefficiencies people have normalized

Best practices:

• Get permission but minimize disruption
• Follow specific work items through the process
• Note what you see without judgment
• Observe at different times (shifts, days, seasons)

Electronic Data Capture

Modern systems can observe processes automatically.

Real-time Instrumented Data

Systems that capture events as they happen:

Advantages:

• Continuous coverage without observer fatigue
• Objective measurements
• Large sample sizes

Challenges:

• May not capture context or reasons
• Data quality issues
• Requires technical access

Historical Data Analysis

Mining past records for process insights:

• Database records - Transaction history, status changes
• Log files - System events, error patterns
• Report archives - Performance metrics over time
• Maintenance records - Failure patterns, response times

Example:

"We pulled three years of flight, maintenance, and supply data from military databases. The historical analysis revealed patterns that weren't visible in day-to-day operations—certain part failures clustered around specific missions and environmental conditions."

Visual and In-Person Methods

Note-Taking Observation

The most basic approach: watch and record.

Tools:

• Paper notebooks (always works)
• Tablets for structured data entry
• Voice recordings (with permission)
• Sketching for spatial layouts

Structured observation forms help consistency:

Checksheets and Tally Sheets

Pre-formatted forms for counting events:

Defect Type Tally (Shift: Day | Date: _______)

Scratch marks: |||| |||| ||
Dents:         |||| |
Missing parts: |||
Wrong color:   ||
Other:         ||||

Video and Photography

Recording for later analysis is powerful when:

• Activities happen too fast to observe in real-time
• Multiple people work simultaneously
• You need to analyze the same sequence multiple ways
• You want to share observations with others

Best practices:

• Get explicit permission
• Explain purpose to reduce self-consciousness
• Position cameras to capture relevant action
• Plan for storage and review time

Post-observation analysis:

1. Review footage at various speeds
2. Timestamp key events
3. Create time-motion studies
4. Share clips to validate interpretations

Interview-Based Observation

Sometimes you can't watch directly. Structured interviews become a form of observation.

Effective Interview Techniques

Question types:

Validation approach:

• Take notes during interview
• Summarize understanding back to SME
• Ask them to correct or confirm
• Follow up on ambiguities

Focus Groups and Workshops

When you need multiple perspectives at once:

• JAD Sessions (Joint Application Design) - Structured workshops with stakeholders
• Process mapping workshops - Group creates process flow together
• Kaizen events - Intensive improvement workshops

Benefits:

• Multiple viewpoints surface quickly
• Disagreements reveal process variations
• Builds buy-in for changes

Risks:

• Dominant voices may overshadow others
• Group dynamics affect what's shared
• Harder to capture everything

Documentation-Based Observation

Documents serve as artifacts of past process behavior.

What Documents Reveal

Reading Between the Lines

Documents tell you what's supposed to happen. Look for clues about what actually happens:

• Handwritten notes on forms suggest missing fields
• Frequent exceptions in logs indicate unrealistic rules
• Workaround documentation shows where official process fails
• Version history reveals how understanding has evolved

External Observation Sources

Sometimes observation extends beyond your organization.

Secondary Research

• Industry benchmarks
• Academic studies
• Competitor analysis
• Regulatory guidance

Site Imagery

"Google Earth and satellite imagery helped us understand the physical layout of remote facilities before site visits. We could see loading dock configurations, parking patterns, and traffic flow without traveling."

Public Information

• Annual reports
• Press releases
• Job postings (reveal internal processes)
• Customer reviews

Practical Observation Tips

Managing Observer Effect

People behave differently when watched. Minimize this by:

• Explaining your purpose - Reduce anxiety about being evaluated
• Observing longer - People revert to normal behavior over time
• Being unobtrusive - Don't hover or interrupt
• Multiple observations - Compare different times and observers

Handling Sensitive Observations

When you observe problems:

• Don't blame individuals - Focus on the process, not people
• Consider context - Workarounds often exist for good reasons
• Verify patterns - One observation isn't a trend
• Report constructively - Describe what you saw without judgment

Recording Effectively

Capture enough to be useful later:

• Time stamps - When did things happen?
• Actors - Who was involved (roles, not names for sensitive issues)?
• Actions - What exactly occurred?
• Artifacts - What documents, tools, or systems were used?
• Context - What else was happening?
• Your questions - What didn't you understand?

Combining Observation Methods

No single method captures everything. Effective process analysis combines approaches:

Triangulation

Validate findings by checking multiple sources:

• What documents say should happen
• What interviews say happens
• What observation shows actually happens
• What data indicates has happened

Discrepancies are often more interesting than agreements—they reveal where improvement opportunities hide.

Key Takeaways

• Observation reveals what documentation and interviews miss
• Guided walk-throughs build understanding; unguided observation reveals reality
• Electronic systems can observe continuously what humans can't
• Combine multiple methods to triangulate findings
• Manage observer effect and handle sensitive findings constructively
• Document observations systematically for later analysis