Interaction Logging and Instrumentation Planning

Description

Plans and builds the capture of user interaction data, through event tracking plans, in-app telemetry, proxies or desktop loggers, driven by research questions and privacy limits.

Log analysis is only as good as what was logged, and data cannot be collected retroactively. Instrumentation planning starts from research questions: what you want to measure and why decides which events, at what granularity, to capture. For a web product, the team walks the site, lists every meaningful user action, checks what is already tracked, and defines a category/action/label/value scheme robust enough for future events. Other routes suit other questions: instrumenting application code to log commands (Microsoft's opt-in programme logged billions of Office sessions and guided the ribbon redesign); proxies that log or rewrite traffic across many sites, or inject scripts to record mouse movements; and desktop activity loggers that record keyboard, mouse and window events across applications to study multitasking and interruption. Each needs consent, opt-out and limits on captured text.

Attributes

Loss type
Renamed
Effort
Weeks

Steps

  1. Derive the needed events and granularity from explicit research questions, stating for each what is captured and why.
  2. Choose the capture route: analytics event tracking, in-app instrumentation, a logging proxy or script injection, or a desktop activity logger.
  3. For web and app products, walk the interface, list every meaningful user action, and audit which are already tracked.
  4. Define a naming taxonomy (for example category, action, label, value) and tag events by category and granularity so analyses can filter.
  5. Build in consent, opt-out and privacy limits: capture that keys were pressed rather than which, truncate titles, exclude file names and addresses.
  6. Have developers implement, then verify data on test sessions, including crash recovery and dropped events.
  7. Parse, aggregate and visualise logs as counts, rates, sequences and paths, and revisit the plan as questions change.

Sources

  • Research Methods in HCI (Lazar et al., 2017)
  • Practical Web Analytics for User Experience (Beasley, 2013)
  • Product Management for UX People (Crumlish, 2022)

Also known as

  • Proxy-based web interaction logging
  • Web proxy logging
  • WebQuilt
  • JavaScript page instrumentation
  • Desktop activity and keystroke logging
  • Activity logger
  • Instrumented software for feature-usage logging
  • Software instrumentation
  • Usage telemetry
  • Analytics instrumentation planning
  • Event tracking plan
  • Tracking coverage audit

Categories

  • Analytics & Behaviour
  • Product Analytics & Strategy