Pendo installation audit: what it covers and what you get.

Often, after installation, the assumption is that the data is flowing correctly but sometimes isn't. A Pendo installation audit is the process of finding out exactly what is and isn't working, before you build dashboards, onboarding flows, or business decisions on top of a foundation you've never verified.

The cost of skipping this step is rarely visible straight away. Broken tracking doesn't throw errors. It quietly produces numbers that look plausible and are wrong. Teams then spend months debating product decisions backed by inflated adoption figures or sessions that were never recorded. The audit covers four areas.

The snippet needs to be present on every route, firing consistently, and identifying visitors correctly. In React apps, the snippet can end up conditionally rendered meaning entire user segments never load the agent, silently dropping sessions with no alert.

Identity is the other half. Anonymous visitors who should be identified, visitor IDs that change between sessions, and accounts that never get mapped all fragment your data: the same human shows up as three different visitors, and account-level reporting becomes guesswork. The audit verifies the identify calls against how your authentication actually behaves, including edge cases like SSO redirects and embedded views.

Selectors built on positional references like nth-child, or broad attribute selectors, silently break after frontend updates. The audit reviews every tag for selector stability, data integrity, and naming consistency and flags features recording zero data for investigation.

Naming matters more than teams expect. When tags are named after CSS classes or sprint tickets instead of what users actually do, every report needs a translator. Part of the audit output is a naming convention your whole team can read without asking the person who created the tags.

The most critical gap in almost every DIY setup: internal users are not excluded from analytics, inflating every engagement metric the product team reports on. Without clean segments, you can't target guides, filter reports by account type, or build a meaningful onboarding flow.

Segments are only as good as the metadata behind them. The audit checks which visitor and account fields are actually being sent (plan tier, role, signup date, account owner) and which reports and guides depend on fields that are empty or stale. This area is where I find the most issues, by a distance.

Common issues include walkthroughs split across separate guides instead of configured as steps, throttling left off so users see the same guide repeatedly, and CTAs limited to dismiss which produces no measurable engagement signal. A guide that annoys users is worse than no guide, and without engagement CTAs you can't even tell which ones are annoying.

The whole process takes about a week and needs very little of your team's time: read-only admin access and one call at the end.

1. Day 1Access and context: read-only Pendo admin, a quick walkthrough of your product, and your top three questions about the data.
2. Days 2–4Full review of the four areas: snippet and identity, every feature and page tag, segments and metadata, and all live guides.
3. Day 5Written report delivered, then a walkthrough call to go through the findings and agree what gets fixed first.

The output is a structured report covering every finding, its impact on data reliability, and the exact steps to fix it ordered by priority. Most teams go from uncertainty about their Pendo data to a clear, sequenced remediation plan in under a week. A few examples of what findings look like in practice:

Victor did a full audit of our initial Pendo installation and got us up to speed. We are now working on creating guides and a reporting dashboard, and are excited to see the analytics behind the improvements.

Renee C, The PUSH Agency