Platform reference · Pendo · Resource Center · Product training video · WCAG 2.1 AA · Digital adoption

Pendo captions: product training video in the Resource Center, walkthrough recordings, and WCAG 2.1 AA compliance

Pendo is the leading product analytics and digital adoption platform, used by SaaS product teams to guide customers and employees through product workflows with in-app guides, walkthroughs, tooltips, and the Pendo Resource Center. The Pendo Resource Center is a self-service help and training hub embedded directly inside the product — customers and employees access it without leaving the application. It can contain video tutorials, feature walkthroughs, release notes, onboarding checklists, and help documentation. Video content in the Pendo Resource Center is subject to WCAG 2.1 AA SC 1.2.2 caption requirements for prerecorded video. Product training video served through Pendo contains a high-density vocabulary failure surface: product analytics terminology (DAU, MAU, NPS, retention cohort, funnel analysis, feature adoption rate, guide engagement, session replay), proprietary feature names that are invented compound words or portmanteau brand terms, and platform workflow vocabulary that describes specific UI flows and integration patterns. All three vocabulary categories fail in generic automatic speech recognition. ADA Title II applies to public sector customers accessing Pendo-embedded training. Section 508 applies to federal agency customers. EAA applies to EU customers. GlossCap's glossary-biased captioning recovers accurate product vocabulary from the acoustically ambiguous audio of product training recordings.

TL;DR

Pendo Resource Center can embed video tutorials and product walkthrough recordings that customers and employees access directly within the SaaS product. This video is subject to WCAG 2.1 AA caption requirements. Product analytics vocabulary (DAU, MAU, NPS, cohort analysis), proprietary feature names, and workflow terminology fail in generic STT. Pendo video content typically lives on external hosting (YouTube, Vimeo, Wistia, Loom) and is embedded in the Resource Center — captions must be managed at the hosting platform level. GlossCap prepares corrected VTT files for these videos using glossary-biased captioning, which are then uploaded to the hosting platform before embedding in Pendo.

Pendo as a product training platform: the Resource Center and video captioning scope

What Pendo does and where video appears

Pendo was founded in 2013 and has grown into the dominant product analytics and in-app guidance platform for SaaS companies. Its core capabilities are: product usage analytics (feature adoption rates, user journey paths, funnel analysis, retention cohorts, NPS surveys), in-app guide authoring (walkthroughs, tooltips, modals, banners, checklists), and the Resource Center — an in-app self-service hub that customers access directly within the product without switching to a separate knowledge base.

From a captioning perspective, the Pendo Resource Center is the primary video delivery surface. Resource Center modules can contain:

Video tutorials — screen recordings or talking-head recordings explaining how to use specific product features. These are typically 2–8 minutes and are hosted on YouTube, Vimeo, Wistia, or Loom and embedded in the Resource Center via iframe or embed code.
Product walkthrough recordings — recordings of the in-app walkthroughs themselves, for customers who prefer video documentation over live guide interactions. These are often produced by customer success teams or product teams and hosted externally before embedding.
Release note videos — "what's new" video summaries of product releases, produced by product marketing teams. Duration: 2–5 minutes. These contain the highest density of new terminology, since every release introduces new feature names that are not yet in any STT model's corpus.
Onboarding video sequences — structured video onboarding sequences (similar to an LMS onboarding curriculum but embedded in the product) for new users. These are mandatory for new customer onboarding flows that use Pendo as the delivery mechanism.

Pendo also enables product teams to record guide previews and walkthrough demonstrations outside the Resource Center context — these recordings are used in customer success emails, sales demos, and internal product documentation, creating additional captioning surfaces beyond the in-app Resource Center.

Where Pendo video is hosted and why caption management is at the hosting layer

Pendo itself does not host video. Video content embedded in the Pendo Resource Center is hosted on external platforms — most commonly YouTube, Vimeo, Wistia, or Loom — and embedded in the Resource Center module via the hosting platform's embed code or iframe. This architectural choice (Pendo embeds video; it does not store video) has a direct implication for captioning: captions must be managed at the hosting platform, not within Pendo itself.

A video tutorial embedded in the Pendo Resource Center displays whatever caption track the hosting platform serves:

For YouTube-hosted video, the caption track is YouTube's auto-generated track unless a corrected VTT or SRT file has been uploaded to YouTube Studio by the content owner.
For Vimeo-hosted video, the caption track is Vimeo's auto-generated track or an uploaded SRT/VTT if the content owner has added one in Vimeo's caption manager.
For Wistia-hosted video, Wistia supports SRT caption file upload through the video manager. The Wistia player displays the uploaded SRT as a selectable caption track.
For Loom-hosted video, Loom auto-generates captions for all recorded videos. These auto-captions fail on the same product vocabulary as all generic STT — Loom's auto-captions are not equivalent to corrected, glossary-biased captions.

To caption Pendo Resource Center video correctly, the process is: (1) identify all video resources embedded in the Resource Center, (2) obtain the source video files for each, (3) prepare corrected VTT/SRT files using GlossCap's glossary-biased captioning with the product vocabulary glossary, (4) upload the corrected caption files to the hosting platform (YouTube Studio, Vimeo caption manager, Wistia video manager), (5) verify the corrected captions are served by the hosting platform before the Resource Center embeds them. The Resource Center embed itself does not need to change — the caption fix happens at the hosting layer.

Internal employee training delivered through Pendo

Beyond customer-facing Resource Center content, some organizations also use Pendo for internal employee training — particularly for rolling out new software to internal employees and guiding them through new workflows. In this internal-training use case, Pendo functions similarly to a digital adoption platform (DAP): it delivers guided walkthroughs, embedded training videos, and self-service help to internal employees using the software. Internal employee training delivered through Pendo is subject to ADA Title I accommodation requirements for hearing-impaired employees, the same as any other mandatory employee training platform.

The vocabulary failure surface in Pendo training video

Product analytics terminology

Pendo users — product managers, customer success managers, and growth teams — speak a dense product analytics vocabulary in training video. This vocabulary is a mix of general SaaS analytics terms (which generic STT handles inconsistently) and Pendo-specific analytics terminology (which generic STT handles poorly):

General SaaS analytics vocabulary: DAU (Daily Active Users), WAU (Weekly Active Users), MAU (Monthly Active Users), DAU/MAU ratio, retention rate, churn rate, NPS (Net Promoter Score), CSAT, time-to-value, onboarding completion rate, feature adoption rate, product-led growth (PLG), activation rate, expansion revenue, logo retention, net revenue retention (NRR), gross revenue retention (GRR). These acronyms transcribe inconsistently — DAU may become "do," MAU may become "mow," NPS may become "ends" depending on pronunciation speed.
Pendo-specific analytics terminology: Page view, feature event, track event, guide interaction, guide engagement rate, poll response rate, NPS response rate, path analysis, funnel step, retention cohort, event tagging, data explorer, dashboard widget, integration data. "Guide interaction" becomes "guide action" or "guide fraction" in noisy recordings. "Track event" may become "tracked event" or "tracking event." "Funnel step" is transcribed correctly but may be confused with "final step" in fast speech.
Pendo product names and features: Pendo Feedback (formerly Receptive), Pendo Adopt, Pendo Engage, Pendo Analyze, In-app Guide Builder, Resource Center, Segment builder, Integration hub, Data Sync, Journey builder. These proprietary Pendo product names fail in generic STT when the training content is produced by narrators speaking quickly through feature lists.

Product vocabulary in product knowledge training

Video tutorials in the Pendo Resource Center are explaining the features of the product that Pendo is deployed on — not Pendo itself. A SaaS company that uses Pendo for customer training is delivering video tutorials about their own product (for example, a project management SaaS explaining how to use their Gantt chart feature, resource allocation module, or API integration). This product vocabulary is entirely specific to the SaaS company's product and is not in any generic STT model's corpus.

This creates the highest-density vocabulary failure surface in the Pendo captioning context: the video describes specific UI workflows and feature names that are unique to the product, narrated at a pace optimized for a viewer following along in the product, with screen share audio that may contain ambient keyboard and mouse sounds. The combination of product-specific vocabulary, screen share audio quality, and follow-along pacing creates a caption failure rate that generic STT cannot adequately address.

Workflow and integration vocabulary

Product training video for Pendo-using companies often covers integration workflows — how the product connects with CRM, ERP, analytics, or data warehouse tools. Integration workflow vocabulary includes both the integration tool names and the specific integration configuration terminology:

Salesforce integration vocabulary: Object, Record Type, Custom Field, SOQL, API key, Apex trigger, Field mapping
Slack integration vocabulary: Slash command, Webhook, Channel notification, App manifest, OAuth scope
Analytics platform vocabulary: Segment event, Mixpanel property, Amplitude cohort, BigQuery dataset, Snowflake schema

These integration vocabulary terms appear in product training video whenever the feature being demonstrated has integration dependencies. The correct captioning of integration vocabulary requires a glossary that includes both the integration tool names and the specific configuration terminology used in the narration.

Compliance requirements for Pendo Resource Center video

WCAG 2.1 AA SC 1.2.2 for all prerecorded Resource Center video

WCAG 2.1 AA SC 1.2.2 requires synchronized captions for all prerecorded video with audio content that is published on the web or in web-based applications. The Pendo Resource Center is a web-based application module — video content embedded in it is subject to the same WCAG caption standard as video on a publicly accessible website. This applies whether the Resource Center is customer-facing or employee-facing.

The 99%+ accuracy standard that WCAG compliance programs require means that auto-generated captions from YouTube, Vimeo, Loom, or Wistia are not sufficient for most product training video — particularly video with product-specific vocabulary. A product tutorial video that uses the product's feature names, workflow names, and integration terminology in every sentence will have auto-caption accuracy well below 99% for the product-specific vocabulary portions of the narration.

ADA Title II for public sector customers

SaaS companies whose product is used by state and local government agencies, public universities, or other public entities are serving customers who have ADA Title II obligations. ADA Title II requires that digital services provided by or used by public entities be accessible. If a public entity uses a SaaS product with a Pendo Resource Center for employee training, the video content in that Resource Center is covered by the public entity's ADA Title II obligations. SaaS companies serving public sector customers may face contractual WCAG requirements from their public sector customer agreements that require the product itself (including in-app training content) to meet WCAG 2.1 AA.

The ADA Title II deadline for many state/local government entities was April 24, 2026 — making the compliance window active now. Public sector SaaS customers are actively auditing their vendors' products for WCAG compliance, including in-app help and training content. A Pendo Resource Center with uncaptioned or auto-captioned video may appear in a public sector customer audit finding, creating remediation pressure on the SaaS vendor. See ADA Title II captions for the complete compliance framework.

Section 508 for federal agency customers

Federal agencies subject to Section 508 require that any information and communication technology (ICT) they procure or use meets Section 508 standards. Section 508's video caption requirements follow WCAG 2.1 AA. Federal agency customers using a SaaS product with Pendo-embedded training video will audit the training content for Section 508 compliance during procurement evaluations and ongoing contract reviews. Uncaptioned or auto-captioned training video in the Pendo Resource Center is a Section 508 finding that can block federal contract renewal. See Section 508 captions for the federal procurement compliance analysis.

European Accessibility Act for EU customers

SaaS companies with EU customers are subject to the European Accessibility Act (EAA), which has been enforceable since June 2025. The EAA requires that digital products and services sold in the EU meet WCAG 2.1 AA accessibility standards. EU-based customers accessing a Pendo Resource Center with video content are covered users under the EAA. Product training video in the Pendo Resource Center that is served to EU customers must meet the same caption quality standard as WCAG 2.1 AA — auto-generated captions are not sufficient. See EAA captions requirements for the EU compliance analysis.

Caption workflow for Pendo Resource Center video

Step 1: audit the Resource Center video inventory

Begin with a complete audit of all video content in the Pendo Resource Center. Log each video by: hosting platform (YouTube, Vimeo, Wistia, Loom), video URL or embed code, current caption status (no captions / auto-captions / corrected captions), narration language, estimated duration, and content category (onboarding tutorial / feature walkthrough / release notes / integration guide). This audit becomes the captioning project workbook.

Step 2: prepare the product vocabulary glossary

Before captioning any videos, build the product vocabulary glossary that covers the full training content scope. The glossary should include:

Product feature names — every named feature in the product that appears in training narration
Workflow and UI element names — named workflows, named UI sections, named settings, named export formats
Integration partner names — every integration partner whose name appears in integration tutorial narration
Technical terms — API terminology, webhook terminology, data format names, authentication method names that appear in technical tutorial narration
Analytics vocabulary — if the product includes analytics features, add the analytics terminology specific to the product

Step 3: obtain source video files and caption with GlossCap

Download or obtain the source video files for each Resource Center video from the hosting platform. Submit each video to GlossCap with the product vocabulary glossary. GlossCap's glossary-biased captioning engine produces corrected VTT files for each video, with product-specific vocabulary correctly preserved throughout the transcript.

Step 4: upload corrected captions to the hosting platform

Upload the corrected VTT/SRT files to the appropriate hosting platform for each video:

YouTube Studio: Video → Subtitles → Upload file → VTT. See YouTube caption upload guidance for the step-by-step workflow.
Vimeo: Video Manager → Video → Distribution → Subtitles. Upload the SRT/VTT file and set the language. The Vimeo player will display the uploaded caption track. See Vimeo captions for training videos.
Wistia: Video → Customize → Captions → Upload SRT. Wistia supports SRT upload and auto-displays the caption track. See Wistia captions.
Loom: Loom auto-generates captions for all videos and supports replacing them with an uploaded transcript. Use Loom's transcript editor to paste the corrected transcript, or replace the auto-captions with the corrected version.

Step 5: verify the Resource Center embeds display corrected captions

After uploading corrected captions to the hosting platform, verify the Pendo Resource Center embeds display the corrected captions by accessing the Resource Center in the product, opening each video, enabling the CC option, and confirming the corrected caption text appears. In some hosting configurations, a CDN cache flush or brief propagation delay may be required before corrected captions appear in embedded contexts.

See GlossCap pricing

FAQ — Pendo Resource Center captions

Does Pendo itself caption video content in the Resource Center?

Pendo does not caption video content. Pendo is a product analytics and in-app guidance platform, not a video hosting platform. Video content in the Pendo Resource Center is hosted on external platforms (YouTube, Vimeo, Wistia, Loom) and embedded in the Resource Center module via embed code. The caption track for embedded video comes from the hosting platform — not from Pendo. To caption Pendo Resource Center video, you must upload corrected caption files to the hosting platform (YouTube Studio, Vimeo caption manager, Wistia video manager, Loom transcript editor). Once corrected captions are live at the hosting platform, the Resource Center embed displays the corrected caption track automatically without any change to the Pendo Resource Center configuration.

Our Resource Center has 50 videos with auto-generated YouTube captions. How do we assess the caption accuracy problem?

Auto-generated YouTube captions for product tutorial video typically have accuracy in the 80–90% range for common English narration, dropping to 60–80% for product-specific vocabulary sections (feature names, workflow names, integration terminology). The practical approach is to spot-check a sample of 5–10 videos by enabling captions in YouTube and scanning the transcript for product-specific terms. Look specifically for: product feature names that appear as common English words, integration partner names that are mis-transcribed, workflow names that lose their hyphenation or capitalization, and technical terms (API, webhook, OAuth, SAML, JSON, CSV, VTT) that are mis-transcribed in fast narration. If the spot-check shows consistent failures on product vocabulary, the full 50-video library needs corrected captions. The GlossCap batch workflow — submit all 50 videos with the product vocabulary glossary, receive 50 corrected VTT files, upload in a batch to YouTube Studio — is the practical remediation approach for a library of that size.

We use Pendo for internal employee onboarding to our enterprise software. Does ADA Title I apply?

Yes. If you use Pendo to deliver video-based training to internal employees as part of mandatory onboarding or required software training, ADA Title I applies to that training. ADA Title I requires employers with 15 or more employees to provide reasonable accommodations for hearing-impaired employees, including accessible training video. Internal Pendo Resource Center training — walkthroughs, feature tutorials, process documentation videos — is employer-provided training content, not a consumer product. Hearing-impaired employees have the right to accessible captions on mandatory internal training video regardless of the delivery platform. The captioning workflow is the same as for customer-facing content: caption the video at the hosting platform, verify the corrected captions appear in the internal Pendo Resource Center.

How does captioning Pendo video differ from captioning Intercom or Zendesk help center video?

The captioning workflow for Pendo Resource Center video is similar to captioning video in Intercom or Zendesk help centers — all three platforms embed externally hosted video rather than hosting video natively, so captions are managed at the hosting platform (YouTube, Vimeo, Wistia, Loom) rather than within the help/guidance platform itself. The primary difference is the audience and content type. Intercom and Zendesk help center video tends to be general feature documentation with lower specialized vocabulary density. Pendo Resource Center video, especially for product analytics-intensive SaaS companies, contains denser product analytics vocabulary (DAU, MAU, NPS, cohort, funnel, retention) and tighter integration with the in-app user journey — customers access Pendo Resource Center while actively using the product, making vocabulary accuracy even more critical. A user who reads "mow" (MAU misheard) in a caption track while trying to understand their analytics dashboard receives incorrect information at the exact moment of product use.

We're building a Pendo Resource Center from scratch. How do we set up captions correctly from the start?

Build the product vocabulary glossary before producing any training video. A glossary-first approach is significantly more efficient than remediating uncaptioned video retroactively. The glossary should be maintained as a shared document (Notion, Confluence, Google Docs — whichever your team uses for product documentation) and updated with each product release. When a new feature is shipped, add its name, any sub-feature names, and any workflow names to the glossary in the same sprint. When producing video tutorials for the Resource Center: (1) produce the video, (2) submit to GlossCap with the current glossary, (3) receive the corrected VTT, (4) upload to the hosting platform, (5) embed in the Resource Center. This pipeline — produce, caption, upload, embed — takes 24–48 hours total and ensures every Resource Center video has corrected captions from first publication. Starting with accurate captions eliminates the remediation project that organizations face when they launch with auto-captions and have to retroactively correct years of accumulated video content.