Pendo — Functional Teardown

URL: pendo.io Live AI agent name: Leo (Pendo's intelligence layer, evolved from "Agent Mode" announced Autumn 2025), plus Signals (proactive monitoring) and Listen Feedback Agent (feedback-side conversational AI). Threat level: Low (enterprise positioning, opaque pricing well above our ICP), but the patterns are highly mature and worth studying. Pricing entry: Free tier (500 MAU), then paid plans quoted by MAU. Real-world deal sizes report $7K–$200K+/yr. Sources reviewed: homepage, /products, /product/leo, /product/analytics, /product/listen, /product/session-replay, /pricing, /pendo-blog/meet-leo, /pendo-blog/agent-mode-pendo-experience, support.pendo.io docs for Agent Mode, Listen Explore, Signals, Dashboards, Funnels, Guides, NPS, Roadmaps.

1. What it is

Pendo positions itself as an AI-powered analytics and digital adoption platform that brings product analytics, in-app guidance, session replay, user feedback, and roadmapping under one roof. The 2025 Autumn release reframed the entire suite around an "intelligent systems" narrative: every module is now wired into a single AI layer called Leo.

The pitch on the homepage today is two-sided. For the product team, Pendo says, "stop opening five tools to understand one drop in adoption." For the executive, Pendo says, "your product intelligence will come to you in Slack, you don't have to log in to find it." Both sides depend on Leo monitoring continuously and surfacing findings proactively.

Target user (canonical): a Product, CS, or Adoption leader at a 200 to 10,000+ employee company that already has a deployed application, an engineering team to install the snippet, and budget for a multi-five-figure annual contract. Pendo has a Free tier for very small teams, but the gravity of the company is enterprise.

2. How data enters the product

Intake model: passive event auto-capture via SDK snippet, plus integrations.

Once the Pendo JavaScript snippet (or mobile SDK) is installed, Pendo automatically captures every click, page view, and feature interaction on the customer's web or native app. The marketing copy is explicit: "no manual event tagging needed." Teams then visually "tag" Pages, Features, and Track Events inside Pendo's Visual Design Studio to give the raw click stream meaningful names.

Other intake paths:

Web and Mobile SDKs (iOS, Android, React Native, hybrid frameworks)
Data integrations with Salesforce, Zendesk, HubSpot for blending CRM/support data with product events (Listen surfaces this most clearly)
Pendo MCP server (newer, 2026) lets external AI tools query Pendo's data via Model Context Protocol, useful for piping insights into Claude, Cursor, or custom internal agents
Webhooks and API access for enterprise tiers
Feedback intake is its own ingestion lane: users submit requests via in-app widgets, and Pendo also pulls feedback from Salesforce, Zendesk, and similar systems into Listen

Critically, there is no CSV upload, no questionnaire, no "tell us about your product" intake. Pendo assumes you have an instrumented application before day one.

3. Product architecture

Pendo's left sidebar is one of the richer ones in the analytics space. The 2025 redesign organized it into two opinionated groups: Learn (sense-making) and Act (intervention).

Home                          ← personalized landing page, default dashboard
Dashboards                    ← custom + shared boards built from widgets

LEARN
  Visitors                    ← individual users, attributes, history
  Accounts                    ← collections of visitors (companies, workspaces)
  Pages                       ← tagged URLs / route patterns
  Features                    ← tagged UI elements within pages
  Track Events                ← server-side or custom-emitted events
  Product Areas               ← groupings that span Pages/Features/Events
  Reports                     ← Funnels, Paths, Retention, Data Explorer
  Listen                      ← Feedback, Insights, Ideas, Roadmaps, Explore
  Replay                      ← Session Replay library
  Sentiment                   ← NPS, CSAT, PMF, UX-lite surveys

ACT
  Guides                      ← in-app tooltips, walkthroughs, banners, lightboxes
  Orchestrate                 ← multi-channel journeys (in-app + email)
  Roadmaps                    ← Now/Next/Later or Timeline planning
  Resource Center             ← persistent in-app help widget

Agent Analytics               ← (newer module) analytics for AI agents you've shipped

A persistent "Create new" shortcut at the top of the sidebar opens a quick-create menu (dashboard, segment, report, guide, journey, survey, feedback item, tag). "Go to page" via Cmd+K or Ctrl+K is a global search/jump command, mirroring the keyboard-driven navigation common in modern dev tools.

The two-group split (Learn vs. Act) reads as Pendo's own version of the UNDERSTAND → DECIDE → ACT loop that Onboardics and Userpilot both use. Pendo puts the DECIDE layer (Leo) across the sidebar rather than at a single location, so any data surface can pop an "Ask Leo" panel without a context switch.

4. Key screens

Because Pendo does not allow self-serve trials at the level required to capture screens, and because their public docs (support.pendo.io, Zendesk-hosted) block direct HTTP retrieval with 403 in this environment, the screens below are pulled from Pendo's own marketing pages (pendo.io product pages, served via cdn.builder.io). These are the highest-resolution renders Pendo publishes itself and match what is documented on support.pendo.io.

4.1 Home / personalized landing

Pendo's Home is a configurable landing page that defaults to a dashboard of the user's choice. The redesign added widgets pinned by role (PM, CS, exec) and the option to launch directly into Ask Leo from the top of the page. The "Pendo 101" template dashboard is a common default and includes widgets for total active visitors, top features by usage, NPS trend, and a feature-adoption funnel.

4.2 Dashboards

A grid layout of resizable widgets. Widget types include: metric tiles (single number plus variance), line charts (time-series of usage), bar charts (rankings of features or pages), funnels (multi-step conversion), retention heatmaps, NPS trend, and lists of recent feedback items. Dashboards can be shared org-wide, filtered by segment, embedded into Slack via Signals, and exported as PDF.

4.3 Reports: Funnels

Pendo's funnel report is the canonical "did users get through these steps" view. Each step is a tagged Page, Feature, or Track Event. The drop-off per step is shown as a percentage and as an absolute count. A 2026 update added support for AI agent events as funnel steps, so teams can measure whether an AI-assisted action lifted conversion vs. a manual one. The funnel can be filtered by segment, account, and time window, and the chart can be inverted to show "completed" instead of "started" cohorts.

Funnel + Paths + Workflow Journeys — Pendo's user-behavior report stack

4.4 Reports: Paths

Paths render the directed graph of what users did before or after a given Page or Feature. The default view is a horizontal Sankey-style flow with stream widths proportional to user counts and stream colors per outcome. Paths is the report most often used for "where do users go after they hit the dashboard for the first time" exploration.

Understand user behaviors — Pendo Analytics canonical report view

Spot trends across apps — multi-app comparison view

Integrate business data — blending Pendo events with CRM/support data

4.5 Reports: Retention

Standard cohort retention heatmap with rows as cohorts (signup week or first-feature-use week) and columns as relative time periods. A 2025 update added the option to measure retention at the account level rather than the visitor level, which matters for B2B customers where one logo equals many seats. Color scale from white to deep blue with explicit percentage labels per cell.

4.6 Visitors and Accounts

Two parallel directory views. Visitors is a sortable list of every individual user with columns for last seen, total events, NPS score, and any custom metadata fed in via API. Accounts is the same view aggregated by company. Clicking through to a single visitor opens a profile with a timeline of every captured event, a list of guides shown, a list of replays available, and the metadata sidecar. This is the "show me this customer's full story" surface that CS teams use during account reviews.

4.7 Pages and Features (Visual Design Studio)

The tagging surface. Admins install the Visual Design Studio browser extension, open it on top of their live application, and click on UI elements to tag them as Features. Each tag is saved with an auto-captured screenshot for visual confirmation. Tagged Pages and Features then appear as first-class objects everywhere else in Pendo (reports, funnels, dashboards, Replay search). This is the friction point at install time and where many teams stall, which is part of why Pendo's onboarding is now AI-assisted.

4.8 Leo / Ask Leo (conversational AI)

Leo is positioned as an inline AI panel rather than a separate tab. The user clicks "Ask Leo" from any data surface, asks a natural-language question like "Which features are most used by new enterprise visitors this month?", and Leo returns a written response plus a chart pinned to the actual Pendo data the answer relied on. The panel supports @ mentions to scope to specific Product Areas, follow-up turns in the same conversation, and a "save as tracked issue" action that turns a one-off finding into a monitored metric. Under the hood, Leo runs on OpenAI and Google Gemini models behind a Pendo orchestration layer.

Leo — take action without leaving the conversation (Skills picker)

4.9 Signals (proactive monitoring) and Slack integration

Signals is the always-on monitoring component of Leo. It watches the entire data set without being asked: which features are spiking, which accounts have gone quiet, which segments are slipping on retention. When something crosses Leo's threshold, the Signal is delivered into a Slack channel as a card with a one-line headline, a sparkline, and a link back into Pendo. The Slack delivery is the headline UX choice — Pendo is explicit that "your team should not have to log in to Pendo to learn what changed."

Continuous monitoring and proactive Signals — Slack-delivered card and in-app feed

4.10 Listen Explore (feedback AI)

Listen Explore is the natural-language interface over feedback. The user opens Listen, switches to the Explore tab, and either picks a prompt template ("What are the top three complaints from churned customers this quarter?") or writes a free-form prompt. Pendo runs the query across all feedback items in Listen (in-app submissions, Salesforce sync, Zendesk sync), clusters them into AI-generated topics, and renders results as: a metrics overview at the top, a Topics tab showing the top five themes with trend lines and supporting quotes, and a flat feedback-item table for drill-down. Templates are pinned above the input field, similar to the Spark pattern in Mixpanel.

Listen — centralize feedback across in-app, CRM, and support sources

Listen — prioritize ideas by revenue impact

Listen — close the loop with users who submitted feedback

4.11 Listen Feedback Agent

A conversational evolution of Listen Explore. Open Beta as of 2026, surfaced as "Feedback Agent" in the Listen dashboard. The agent maintains conversation state, can synthesize across feedback sources (Pendo in-app submissions, Salesforce cases, Zendesk tickets), and produces data-backed summaries with source quotes. Pitched as the assistant a PM brings to a roadmap meeting instead of an unstructured pile of anecdotes.

4.12 Guides and AI Guide Builder

Guides is the engagement layer. Tooltips, lightboxes, banners, walkthroughs, embedded inline content. Built in the Visual Design Studio (point and click) or in the older Adopt Studio. The 2025 addition was AI Guide Builder: type a natural-language instruction like "Make a three-step walkthrough for new admins to find the billing settings," and Leo drafts the guide content, targeting, and order. The user reviews, edits, and publishes. This mirrors Userpilot's "Create With Lia" handoff almost exactly, but Pendo extends it deeper into the layout and targeting decisions, not just the copy.

Guides — add guides directly into the live interface

Guides — personalize every experience by segment/role

Guides — branching logic for multi-path walkthroughs

AI Guide Builder — jumpstart guide creation from natural-language prompt

4.13 Session Replay

A library of recorded user sessions with filters across visitor attributes, segment, page, frustration signal (rage clicks, error clicks), and date. Each replay opens with a synchronized timeline of the user's clicks, network requests, and console errors. The most operationally useful surface here is the Frustration column, which flags replays where rage clicks or error clicks occurred so a PM can jump straight to the broken parts of the product.

Session Replay — library with Frustration signals (rage clicks, error clicks)

Session Replay — replays across third-party apps

Session Replay — spot, report, and resolve workflow

Session Replay — customer privacy controls (auto-masking)

4.14 Agent Analytics

A newer module dedicated to measuring AI agents that the customer has shipped inside their own product. Tracks interaction counts, abandonment rates, drop-off points, sentiment per turn, and ROI. Notable as a category bet: Pendo is positioning to become the analytics layer for "your AI features," not just your product features. Includes a "tracked issue" surface where saved findings from Ask Leo can be persisted as monitored problems.

4b. Note on real product screens

The images/ folder for this teardown contains 18 real product UI renders pulled directly from pendo.io product pages (Leo, Analytics, Listen, In-App Guides, Session Replay). These are the same renders Pendo itself uses to demonstrate the product, served via their cdn.builder.io CDN — so they are not marketing illustrations, they are real Pendo UI captured by Pendo.

Coverage by surface:

Leo / Ask Leo conversational panel — 02_leo_ask_anything_natural_language.png, 03_leo_take_action_in_conversation.png
Signals (proactive monitoring + Slack delivery) — 01_leo_continuous_monitoring_signals.png
Listen (feedback intake, prioritization, close-the-loop) — 04_listen_centralize_feedback.png, 05_listen_prioritize_ideas_by_revenue.png, 06_listen_close_loop_with_users.png
Analytics (funnels, paths, workflow journeys, multi-app comparison, CRM blend) — 07–10
Guides + AI Guide Builder — 11–14
Session Replay (Frustration signals, multi-app, privacy) — 15–18

Gaps not covered in images/ (because they live only on the Zendesk-hosted support.pendo.io docs, which returned 403 to programmatic requests in this environment):

Pendo 101 default dashboard — see "Take a tour of Pendo" (support.pendo.io/hc/en-us/articles/20826000368411).
Visual Design Studio tagging view — see "Create a guide in Adopt Studio."
Listen Feedback Agent conversation view — see the "Meet Listen's Feedback Agent" blog post (URL captured in Sources; images embedded inside the blog post HTML were not exposed in the same builder.io CDN pattern).
Pricing page tier table — pendo.io/pricing renders the tier table client-side; capture by hand if needed.

If those four are needed, opening the support article in a logged-in browser and taking an OS screenshot is the fastest path.

5. The core diagnosis / analysis flow

Pendo does not have a single "diagnose my product" flow the way our product does. Instead, Pendo offers two AI loops that together cover the same job:

Loop A — proactive (Signals + Leo):

Pendo is installed and data is flowing.
Leo watches everything continuously. Adoption, retention, funnel steps, NPS, feedback volume per topic, replay frustration signals.
When Leo detects something noteworthy (a feature spiking, an account going quiet, a segment slipping), it fires a Signal.
The Signal arrives in Slack as a card with a headline, a chart preview, and a deep link back to the relevant Pendo view.
The user clicks through, lands on a contextual surface (a specific Page, Feature, Visitor, or Account), and can immediately Ask Leo for the why.
Leo answers in natural language, optionally renders a follow-up chart, and offers a "save as tracked issue" pin to keep watching the problem.
If the diagnosis points to a UX intervention, the user clicks into Guides and either builds the guide manually or asks AI Guide Builder to draft one.

Loop B — reactive (Ask Leo + Listen):

The user has a hunch or a question. "Did the new pricing page change improve conversion?" or "What are paying customers complaining about most this month?"
Ask Leo from any data surface (product-side) or open Listen Explore (feedback-side).
Pick a template or write a prompt. Leo turns the prompt into the right report or feedback aggregation.
Review the result, refine with a follow-up turn, save to a dashboard or roadmap item.
If action is needed, hand off to Guides, Orchestrate, or Roadmaps in the same UI.

The strongest pattern here is that Leo is the same intelligence across both loops. The proactive Signal in Slack and the reactive Ask Leo answer in the dashboard come from one orchestration layer, with the same data scope and conversation memory. That coherence is the single biggest investment Pendo made in 2025.

6. Pricing structure

Pendo is famously opaque on pricing. The pricing page lists tiers but only shows a real number on the Free plan. Everything else is "request a quote" and is priced by MAU.

Tier	Price	MAU	Notable inclusions
Free	$0	500	Product analytics, in-app guides (Pendo-branded), NPS surveys (Pendo-branded), roadmaps (Pendo-branded). No API, no integrations, 1-year retention. If you exceed 500 MAU, reports fall back to a random sample of 500 and guide/survey creation freezes.
Base	Custom (quoted by MAU, ~$7K–$25K/yr typical for 1K–10K MAU)	Tiered	Removes Pendo branding, unlocks integrations, multiple environments, segments, NPS, basic Listen.
Core	Custom (~$25K–$50K/yr typical)	Tiered	Adds session replay, engagement scoring, broader Listen access, advanced segmentation.
Pulse	Custom	Tiered	Heavier Listen and Sentiment usage, more roadmapping power, expanded API.
Ultimate / Portfolio	Custom (~$75K to $200K+/yr)	Multi-app, multi-product portfolios	Full Leo, Signals, Agent Analytics, multiple subscriptions, dedicated CSM, SSO, custom roles, advanced governance.

Notes on the pricing model:

MAU is the primary metering axis. Bundles sold in 1,000-MAU increments. Cost per MAU drops as volume rises.
Vendr marketplace data reports a median annual cost of about $48,500 across 523 purchases, with buyers saving ~12% on average through negotiation.
AI features (Leo, Signals, Listen Explore) are positioned as tier-gated rather than usage-capped (the opposite of Mixpanel Spark's per-query cap). Heavier AI access requires moving up the tier ladder.
A separate Agent Analytics module is sold on top, priced by the number of AI agents being monitored inside the customer's product.

For an ICP at $3K–$50K MRR, even the lowest paid Base tier is roughly equal to a founder's entire monthly revenue. Pendo is structurally not for our customer.

7. vs. our product

Dimension	Pendo	Our product
Data intake	Passive: SDK auto-capture, plus Salesforce, Zendesk, HubSpot integrations	Active: CSV upload (Mode A) or intake questionnaire (Mode B), no SDK
Target user	Product, CS, or Adoption leader at a 200 to 10,000+ employee company	Bootstrapped founder at a 1 to 20 person SaaS, no product team
Diagnosis scope	Where users drop off inside the product, plus feedback themes across CRM and support data	Why business growth has plateaued at the company level
Output	Real-time dashboards, Slack signals, in-app guides, roadmaps, replays	Plateau-type diagnosis plus ranked recommendations, one-shot
Execution layer	Deploys guides, banners, surveys, journeys directly into the customer's product	Recommendations only, we do not deploy anything
AI presence	Leo as a continuous intelligence layer plus Signals for proactive Slack alerts plus Listen Feedback Agent	Diagnostic model run on submitted data, no continuous monitoring
Pricing entry	Free at 500 MAU, but real-world contracts start in the mid four figures and median around $48K/yr	$100 to $250/mo flat
Time to value	Weeks (install SDK, tag Pages and Features, get data flowing)	Minutes (upload CSV or answer the intake)
Buyer psychology	"We have a data team, we want a platform"	"I am alone with my product, I need a diagnosis"

The two products do not compete for the same buyer at all. Pendo wants the multi-product enterprise. We want the founder who would never get past Pendo's tagging step. Where the overlap is conceptual but not commercial: both products promise "tell me what's wrong and what to do about it." Pendo proves the proactive-monitoring + Slack-delivery pattern is achievable. We choose deliberately not to compete on continuous monitoring or in-product execution.

The one place Pendo is genuinely informative for us: their AI architecture choice. Leo is the same intelligence everywhere, not a feature buried in one tab. That coherence is what we should aim for with our diagnosis engine. Whether the surface is the intake review, the diagnosis report, or a future "monitor my plateau" mode, the AI's reasoning style and provenance should feel like one engine, not three bolted-on widgets.

8. Patterns worth borrowing

One AI layer surfaced everywhere, not a separate tab. Pendo's Leo lives inside Pages, Features, Track Events, Visitors, Accounts, Dashboards, and the Homepage. Wherever data is, "Ask Leo" is one click away. We should do the same: our diagnosis assistant should be available from the intake screen, the report, every recommendation card, and any future drill-down, rather than confined to a chat tab.

Proactive delivery into the user's workflow, not a destination they have to visit. Signals into Slack is a strong pattern. We do not need Slack, but the principle holds: when a follow-up insight is ready (new benchmark data, a quarterly re-diagnosis, an industry shift), send it to the founder via email or wherever they already work, not as a notification that requires logging in.

Save findings as tracked issues. Leo lets users pin a finding from a one-off conversation as a persistent monitored issue. We should let founders pin a plateau hypothesis ("I think pricing is the bottleneck") and have our product re-check it over time. This is the cheap version of continuous monitoring without needing an SDK.

Prompt templates above the input field. Listen Explore opens with a grid of prompt templates. Same Spark/Mixpanel pattern. Worth borrowing on our intake screen too, especially in Mode B (questionnaire): six to eight templates that anchor what kind of input is useful.

Two-group sidebar split (Learn vs. Act). The deliberate split into "understand" and "intervene" helps even brand-new users locate where they are mentally. Even if we do not have an Act group, organizing our nav around "see the diagnosis" vs. "build the action plan" would reduce cognitive load.

AI provenance: each AI-generated artifact carries its underlying scope. When Leo answers a question, the response shows which Pages, Features, or Product Areas the answer is scoped to. Our diagnosis should equally show which inputs (the CSV column, the intake answer, the benchmark) drove each finding.

Generated titles for AI outputs. Like Spark, Leo gives generated charts and reports their own titles rather than "Result 1." We should auto-name diagnostic findings.

AI Guide Builder framing for the generative step. Pendo's framing — "describe what you want, Leo drafts it, you review and publish" — is the right way to introduce generative AI for a non-technical user. We should adopt the same three-step framing for any generative step we add (draft → review → ship), including a clear "this was AI-drafted, here is why" provenance line.

A glossary surface (Lexicon-style governance). Pendo's Lexicon names every event and property. Our equivalent could be a small "what each plateau-type means in your language" reference, so a founder doesn't have to interpret jargon.

Founder-quoted blog content as discovery. Pendo's blog has high-quality long-form pieces ("Meet Leo," "Agent Mode," "Rethinking the product roadmap") that position the team as thoughtful operators, not just feature shippers. Worth doing on our side as a category-building exercise once we have early case studies.

Patterns to deliberately NOT borrow

Continuous SDK monitoring. Wrong fit. Our user does not have engineering capacity to install an SDK and our diagnosis does not need real-time event data.
The execution layer (tooltip, modal, walkthrough, journey deployment). That is product-team territory. Our user would not use it, and offering it would dilute our positioning.
Opaque MAU pricing. Pendo's "request a quote" model is brutal for the founder buyer. Our pricing must be visible on the marketing site, predictable, and not require a sales conversation.
The tagging step. Asking a non-technical founder to install a Chrome extension and click every UI element to "tag" it is a non-starter. Our intake (CSV or questionnaire) is deliberately the opposite.
Agent Analytics as a separate product. Pendo monetizes "analytics for AI agents you shipped" as its own SKU. We do not need this category. Our user does not have AI agents in production.
Five-tier paid pricing. Free, Base, Core, Pulse, Ultimate is too many. Three tiers maximum. Decision fatigue is real for founders.

Sources

Marketing site (sections 1, 2, 3, 6):

Homepage: https://www.pendo.io/
Pendo AI / Leo: https://www.pendo.io/ai/ and https://www.pendo.io/product/leo/
Product Analytics: https://www.pendo.io/product/analytics/
In-App Guides: https://www.pendo.io/product/in-app-guides/
Session Replay: https://www.pendo.io/product/session-replay/
Listen: https://www.pendo.io/product/listen/
Roadmaps: https://www.pendo.io/product/roadmap/
Agent Analytics: https://www.pendo.io/product/agent-analytics/
Pricing: https://www.pendo.io/pricing/
Pendo Free: https://www.pendo.io/pendo-free/

Blog and release pages:

Meet Leo (Mar 2026): https://www.pendo.io/pendo-blog/meet-leo/
Autumn 2025 release: https://new.pendo.io/autumn-2025/ and https://www.pendo.io/pendo-blog/the-pendo-autumn-2025-product-release-intelligent-systems/
Agent Mode launch: https://www.pendo.io/pendo-blog/agent-mode-pendo-experience/
Listen Feedback Agent: https://www.pendo.io/pendo-blog/meet-listen-feedback-agent/
Pendo Sentiment: https://www.pendo.io/pendo-blog/meet-pendo-sentiment/
Visual Design Studio: https://www.pendo.io/pendo-blog/introducing-pendo-visual-design-studio-bringing-power-app-guides-everyone/

Public docs (section 4 source, support.pendo.io):

Take a tour of Pendo: https://support.pendo.io/hc/en-us/articles/20826000368411-Take-a-tour-of-Pendo
Dashboards: https://support.pendo.io/hc/en-us/articles/360032203651-Dashboards
Dashboard widgets: https://support.pendo.io/hc/en-us/articles/360032752991-Dashboard-widgets
Funnels: https://support.pendo.io/hc/en-us/articles/360031863292-Funnels
Understand user behavior with reports: https://support.pendo.io/hc/en-us/articles/22172817933595-Understand-user-behavior-with-reports
Analyze product usage with Leo (beta): https://support.pendo.io/hc/en-us/articles/39748859051931-Analyze-product-usage-with-Leo-beta
What's new in AI features: https://support.pendo.io/hc/en-us/articles/44480532063899-What-s-new-in-AI-features
Signals in Pendo and Slack (beta): https://support.pendo.io/hc/en-us/articles/40921734730651-Signals-in-Pendo-and-Slack-beta
Set prediction alerts to Slack: https://support.pendo.io/hc/en-us/articles/43488218914203-Set-prediction-alerts-to-Slack
Overview of Pendo Listen: https://support.pendo.io/hc/en-us/articles/18159674293531-Overview-of-Pendo-Listen
Explore feedback with AI in Listen: https://support.pendo.io/hc/en-us/articles/37717114561819-Explore-feedback-with-AI-in-Listen
Set up Listen Explore: https://support.pendo.io/hc/en-us/articles/37643157438107-Set-up-Listen-Explore
Guides overview: https://support.pendo.io/hc/en-us/articles/27240321140763-Guides-overview
Create a guide in Adopt Studio: https://support.pendo.io/hc/en-us/articles/27777837799067-Create-a-guide-in-Adopt-Studio
Watch replays: https://support.pendo.io/hc/en-us/articles/18047528068123-Watch-replays
Filter the replay library: https://support.pendo.io/hc/en-us/articles/32153640803739-Filter-the-replay-library
Roadmaps in Listen: https://support.pendo.io/hc/en-us/articles/12928869827867-Roadmaps-in-Listen
Net Promoter Score (NPS) in Pendo: https://support.pendo.io/hc/en-us/articles/43643773661083-Net-Promoter-Score-NPS-in-Pendo
Pendo Free: https://support.pendo.io/hc/en-us/articles/6069107464091-Pendo-Free
MAU limits: https://support.pendo.io/hc/en-us/articles/15968607780123-MAU-limits
Overview of Agent Analytics: https://support.pendo.io/hc/en-us/articles/45209760982683-Overview-of-Agent-Analytics

Third-party pricing intelligence:

Vendr marketplace (median deal data): https://www.vendr.com/marketplace/pendo
Costbench: https://costbench.com/software/product-analytics/pendo/
ProductBridge pricing analysis: https://productbridge.io/blog/pendo-pricing
Usercall pricing analysis: https://www.usercall.co/post/pendo-pricing

Image provenance: 18 product UI renders downloaded from pendo.io product pages (Leo, Analytics, Listen, In-App Guides, Session Replay), served by Pendo's own cdn.builder.io CDN. These are the highest-resolution renders Pendo publishes and match the documented UI on support.pendo.io. The four gaps listed in section 4b (Pendo 101 dashboard, Visual Design Studio tagging view, Listen Feedback Agent conversation view, pricing tier table) require a logged-in browser screenshot because support.pendo.io (Zendesk) returned 403 to programmatic requests. Image rescue pass: 2026-05-25.

Channels NOT used: G2 and Capterra were not queried because previous teardowns showed they return bot-protection pages or unrelated content. Pendo's community forum (community.pendo.io) was referenced through search snippets but not crawled directly.

Competitor functional comparison