Events vs Properties vs User Traits

The most common mistake in product instrumentation is conflating three distinct data constructs. Getting this right at setup prevents months of schema cleanup later.

Events are actions a user takes at a specific moment in time. They are verbs: signup_completed, report_exported, payment_initiated. Every event should answer: what happened, when, and who did it.

Properties are metadata attached to a specific event. When a user exports a report, properties might include: report_type: "revenue", format: "PDF", filters_applied: true. Properties transform raw events into analysable data points.

User Traits are persistent attributes of a user that change over time but describe who they are, not what they just did: plan: "Growth", company_size: "51-200", industry: "fintech", country: "India". Traits allow segmentation of any funnel or cohort.

Naming Conventions That Scale

Adopt the Object-Action naming pattern from day 1 — it is the industry standard used by Amplitude, Segment, and Mixpanel internally:

• Format: object_action — e.g. dashboard_viewed, invoice_created, team_member_invited
• Always snake_case, always lowercase — avoid camelCase and spaces
• Be specific about the object — report_exported not export_clicked
• Use past tense for completed actions — payment_completed not payment_complete
• Never embed user IDs, timestamps, or amounts in event names — those belong in properties

What to Instrument on Day 1

Do not try to track everything. Instrument the 10 events that map your core user journey first:

• Acquisition: signup_started, signup_completed, email_verified
• Activation: onboarding_step_completed (with step name as property), aha_moment_reached (your product-specific activation event)
• Core Value Loop: The 2–3 events that represent the primary action users take when they get value from your product
• Engagement: session_started, feature_used (with feature name property)
• Revenue: plan_upgraded, payment_completed, subscription_cancelled

The 5 Metrics Every PM Should Review Daily

A product health dashboard should answer one question in under 60 seconds: Is the product doing better or worse than yesterday? More than 5 daily metrics is a sign of poor prioritisation, not rigour.

Dashboard Infrastructure

For early-stage startups (pre-Series A), keep it simple: a Mixpanel or Amplitude dashboard shared via link with the founding team is sufficient. Post-Series A, invest in a proper data warehouse (BigQuery or Redshift) with a BI layer (Metabase or Looker) so analysts can build bespoke views without engineering dependence.

Activation Funnel

Every product has a critical sequence of steps between signup and first value delivery. For a B2B SaaS product like an HR tool, this might be: Signup → Verify Email → Add Company Details → Invite First Team Member → Run First Payroll. Measure the conversion rate between every consecutive step and the time taken at each step.

The step with the largest drop-off is your highest-leverage optimisation opportunity. If 60% of users drop between "Add Company Details" and "Invite First Team Member", you need to understand why — is the invite flow too complex? Is the value of inviting a colleague not obvious? Use session recordings (PostHog, Hotjar) alongside the funnel data to form hypotheses.

Feature-Specific Funnels

Beyond activation, build funnels for every major feature that has a multi-step flow. A reporting feature funnel might look like: Reports Tab Clicked → Report Type Selected → Filters Applied → Report Generated → Report Exported. Features with low export rates despite high generation rates suggest users are building reports but not finding them useful enough to share — a quality or relevance problem.

Track these funnels by user segment (plan type, company size, industry) to find which segments get value from which features. This directly informs your ICP (Ideal Customer Profile) refinement.

Reading Retention Curves

A cohort retention chart plots groups of users who joined in the same time period (a week, a month) against whether they returned in subsequent periods. The shape of the curve tells you more about product-market fit than almost any other metric.

Curving to a flat line — even at 20% — is a product-market fit signal. It means some users find enough value to keep returning. Continuously declining to zero means you have not yet found a core retained user base. For Indian B2B SaaS startups, aim for a 6-month retention curve that flattens above 30%.

Identifying Sticky Features

Run feature retention analysis: for each major feature, compare the 30-day retention of users who used that feature in their first week versus users who did not. Features that correlate with dramatically higher retention are your sticky features — they should be front-and-centre in onboarding, not buried in a menu.

This is how Slack discovered that teams which sent 2,000 messages retained at nearly 100% — it became the activation benchmark that shaped their entire onboarding. Indian SaaS companies like Freshworks and Zoho have used similar feature-retention correlation analysis to shape product roadmaps.

India's Digital Personal Data Protection Act (DPDP), 2023

India's DPDP Act came into force in 2023 and fundamentally changes how startups must handle user data, including behavioural analytics data. Key obligations for product analytics:

• Consent before collection: You must obtain explicit, informed consent before tracking any personal behavioural data. Generic terms of service acceptance is no longer sufficient — the consent must be specific to data processing purpose
• Purpose limitation: Analytics data collected for product improvement cannot be repurposed for advertising without a separate consent
• Data minimisation: Only collect analytics data that is necessary for the stated purpose — avoid capturing full URLs (which may contain personal info), raw form inputs, or PII in event properties
• Data localisation: For Indian users, significant personal data may need to be stored on servers located in India — check your analytics vendor's data residency options

Anonymisation Best Practices for Product Analytics

The safest approach is to anonymise at the point of collection. Replace user emails and names with anonymised IDs before sending to any third-party analytics tool. Use PostHog's self-hosted option or Amplitude's data residency feature if Indian data localisation is a concern. Strip PII from all event properties — never log email addresses, phone numbers, Aadhaar numbers, or payment card details as event properties.

GDPR Applicability for Indian Startups

If your product serves users in the European Union, GDPR applies regardless of where your company is incorporated. This means you need a cookie consent banner, the ability to export or delete a user's data on request, and documented legal basis for every processing activity including analytics.

Corpfy's Product Analytics dashboard tracks DAU, MAU, stickiness (DAU/MAU ratio), and feature adoption in one unified view. Usage benchmarks by stage are pre-loaded so you can compare your engagement metrics against what is typical at Seed, Series A, and Growth without running a separate analysis or paying for a product analytics tool.

What you get in the dashboard