Use case · Sales enablement
Sales enablement video captions: product names, competitor names, and the searchable-deck workflow
Sales enablement video is half product pitch, half competitive positioning, half new-feature walkthrough — densely packed with the proper nouns the field rep is going to need to say (and spell) correctly on a customer call tomorrow. The product names. The feature codenames. The competitor names you compare against. The pricing tiers. Captioning these with general STT produces a record full of phonetic guesses on exactly the words an enablement archive must preserve. Worse: the captions become the basis for searchable enablement decks (Highspot, Seismic, Gong, Mindtickle, Lessonly), and a wrong product-name spelling means the rep can't find the segment they remember when they search for it three weeks later.
TL;DR
Sales enablement video lives in the proper-noun layer: product names, feature codenames, competitor names, pricing tiers, account-segment labels. General STT writes phonetic guesses; the resulting captions break the searchable-enablement-archive use case as much as they break the WCAG 2.1 AA accessibility floor. GlossCap's glossary pulls from your product taxonomy and competitor list (Notion / Confluence / spreadsheet); the decoder boosts those tokens; the output SRT/VTT preserves the names a rep needs to grep three weeks later.
The proper-noun density of sales enablement video
A 30-minute sales enablement module on a new feature release will typically contain:
- Your product names. The full product (e.g., "Atlas"), the SKU tier ("Atlas Enterprise"), the feature ("Atlas Workflows"), the codename used internally before launch ("Project Polaris").
- Your competitor names. Three to ten depending on the segment. Some are unambiguous ("Salesforce"); some are ambiguous in audio without context ("Apollo" — the people-data company, the marketing automation, or the Greek god?).
- Your customer names. Win-stories invariably reference customer logos. Some are common dictionary words; many are ambiguous in audio.
- Pricing tiers. "Solo $29", "Team $99", "Org $299" — the spoken price needs to render as the right surface form for the rep to quote correctly.
- Account-segment labels. "ICP Tier 1", "POD-2 territory", "AE-led", "PLG-influenced" — internal taxonomy that a generic model has never seen.
None of these are dictionary words. All of them get mangled by general STT.
The searchable-enablement-archive use case
The dominant downstream consumer of sales enablement video captions is not the live viewer — it's the rep three weeks later, in the middle of a customer call, who remembers seeing a relevant segment and needs to find it fast. Modern enablement platforms (Highspot, Seismic, MindTickle, Gong) index the caption text and offer search across the archive. The user journey:
- Rep is on a call. Customer asks about how Atlas Workflows handles approval routing.
- Rep opens Highspot, types "atlas workflows approval routing".
- Search returns segments where the captions contain those words.
- If the captions said "atlas work flows" (split) or "Atlas workflow" (singular) or "Atlas Work Flows" (incorrect casing), search misses or returns wrong segments.
- Rep cannot find the segment. Falls back to making up an answer or asking a colleague on Slack while the customer waits.
This is the practical loss. The accessibility argument is real and matters; the searchable-archive argument is the one that gets enablement leads to greenlight a captioning vendor change.
The competitor-name problem
Competitor names are a special case worth calling out: they appear constantly in sales enablement video, they're rarely covered by general training corpora at the level of detail enablement scripts use, and getting them wrong makes the captions unusable for the rep researching how to position against a specific competitor.
A typical competitive-positioning segment script: "When we run into Verbit, the angle is enterprise pricing — they're $200+ per month and the buyer needs to be at >1000-seat org. When we run into Rev, we differentiate on the glossary feature — Rev does generic captioning, we preserve your jargon." General STT on this segment is reliable on "Verbit" (proper noun, but loaded enough); shaky on "Rev" if the rep speaks it as one syllable; broken on more obscure competitors. A glossary entry per competitor with the canonical spelling fixes this on first export.
The glossary-biased workflow
- Build a sales-enablement glossary. Pull the product taxonomy (every SKU, feature, codename), the competitor list (with canonical spelling and casing), the customer-logo list (for win-stories), and the internal segment labels into a Notion page or Confluence space. Sync to GlossCap, or paste a flat list.
- Caption new enablement modules in batches. Per release cycle, batch all enablement video into one workspace pass. The glossary is shared; the model improves with each correction across batches.
- Reviewable edit UI for the enablement lead. The amber-highlight UI surfaces every glossary-applied term. Enablement leads (or product marketing) can scrub through to confirm; corrections feed the glossary and improve future passes.
- Export to your enablement platform. SRT works on Highspot, Seismic, MindTickle, Gong, Lessonly. VTT works for HTML5 embeds in your own intranet. See our SRT for training videos and VTT for training videos pages.
Compliance side: the accessibility floor still applies
Sales enablement video is internal training; it inherits the same WCAG 2.1 AA accessibility floor that other internal training does, and is in scope under ADA Title II if the company is a state/local government entity (rare for enablement) or under Section 508 for federal contractors. For pure-private SaaS, the legal exposure is lower; the practical exposure is the same — deaf-or-hard-of-hearing reps need captions to do their job, and the best-quality glossary-aware captions also serve the searchable-archive use case the hearing reps depend on.
Related questions
Can the glossary include things we don't say out loud (e.g., internal codenames before launch)?
Yes — glossary entries are matched against the audio. If the speaker says the codename, the decoder preserves it; if the codename never appears in the audio, the entry is inert. Many enablement teams add the public-name + codename pair so segments recorded before vs after launch all surface in search.
What about call-recording captions (Gong, Chorus)?
Gong and Chorus run their own ASR pipelines optimised for call recording. Captioning those in a separate pipeline duplicates work; for enablement video specifically (recorded training, product walkthroughs, competitive briefs), GlossCap is the upstream. For raw call recordings, your call-intelligence vendor's transcription is usually fine for the search use case as long as their model has a workspace-glossary equivalent.
How does the glossary handle competitor names that change frequently (renames, acquisitions)?
Glossary entries are versioned by your team — when a competitor renames or gets acquired, update the entry, and the change applies to new captioning batches. Old captions are not retroactively rewritten (the historical caption is the historical record); the search-archive can be rebuilt as needed by re-captioning the affected modules.
Can different sales segments have different glossaries?
Yes — workspaces support multiple glossaries; a batch can apply a glossary subset. Enterprise-segment competitive briefs apply the enterprise-competitor glossary; SMB-segment briefs apply the SMB-competitor glossary. Same workspace, different glossary scope per batch.