fix(survey): improve feedback survey initialisation and telemetry

[sevoku]

This pull request focuses on improving the survey initialization and telemetry:

• update A/B test selection probability to 50%
• Add more focused telemetry to initSurvey (survey.init event)
  • isCandidate: whether qualified or not
  • reason: string describing the decision step to disqualify as a candidate
    • 02_prompt_version_only_once:
      • skipVersion: opt-out version
    • 03_survey_taken_recently
      • surveyTakenDate: date when the survey was taken
      • daysSinceTaken: days since the survey was taken
    • 04_opted_out_recently
      • optOutDate: date when user opted out
    • 05_prompted_today
      • lastSessionDate: date of the last session
    • 06_insufficient_sessions
      • sessionCount: current session count
    • 07_not_in_ab_test_group
      • acceptedForABTest: whether machine has been selected
      • normalizedValue: normalized score
    • 08_random
      • usedFallbackSelection: whether A/B scoring failed previously
• add telemetry tracking to promptAfterActionEventually with a new survey.measure-score event
  • Properties:
    • triggerAction: the action that triggered the prompt
    • triggerActionScore: the score increase contributed by the action
    • experience: the experience
    • experienceScore: the current experience score
    • fullScore: the current experience score
    • isSurveyDisabledGlobally
    • wasPromptedInSession
    • scoreTargetReached: whether the score target was reached
    • isCandidate: whether qualified or not

Improvements to A/B Test Selection:

• src/utils/survey.ts: Updated the A_B_TEST_SELECTION value from 0.25 to 0.5 for consistent testing.

Enhancements to Telemetry Logging:

• src/utils/survey.ts: Added detailed telemetry properties in the promptAfterActionEventually function to log various survey-related metrics.

Refactoring Candidate Status Logic:

• src/utils/survey.ts: Introduced the setCandidateStatus helper function to streamline setting the candidate status and logging telemetry properties. This function is used in multiple places to simplify the code and ensure consistency. [1] [2] [3] [4] [5] [6] [7]

Consistent Testing:

• src/utils/survey.initSurvey.test.ts: Set SurveyConfig.settings.A_B_TEST_SELECTION to 0.5 within the mockABTest helper function to ensure consistent testing.

[Copilot]

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

Comments suppressed due to low confidence (1)

src/utils/survey.ts:215

• The configuration object is defined directly on SurveyConfig, yet its properties are accessed via SurveyConfig.settings. Ensure a consistent object structure to avoid potential runtime errors.

if (SurveyConfig.settings.DEBUG_ALWAYS_PROMPT) {

[tnaum-ms]

@sevoku I should have checked current PRs :-D

af83ef4

[tnaum-ms]

@sevoku I should have checked current PRs :-D

af83ef4

Please add a "number prefix" to rejection reasons, similar to the file linked. This should make it easier to generate a sorted funnel-chart.