Captioning executive communications video: all-hands recordings, CEO messages, earnings calls, and investor day presentations

Why executive communications is a distinct captioning category

The employee communications captioning guide identifies the structural ownership problem that leaves most employee communications video uncaptioned: L&D owns the caption programme but defines its scope as training content; corporate communications or People/HR owns the content production but has no captioning infrastructure; IT owns the platforms but treats caption accuracy as a content question. Executive communications sits inside that gap and makes it harder in three specific ways.

Vocabulary volatility

Training video vocabulary is stable. A safety compliance module uses the same OSHA terminology for years. A product certification course uses the product vocabulary as of its production date. The L&D caption programme can build a glossary once and maintain it quarterly. Executive communications vocabulary changes every quarter — or every week, depending on what is happening strategically. Q3 earnings call vocabulary includes the Q3 revenue figure, the Q3 non-GAAP EPS, the Q3 guidance for Q4, and the names of whatever new products or initiatives were launched in Q3. None of that vocabulary exists in any glossary before the call. The M&A announcement on Tuesday introduces the target company name, the acquirer's acquisition team, the transaction timeline, and the regulatory bodies expected to review the deal — all new terms, all with specific pronunciation conventions, most not guessable from spelling. The restructuring announcement on Thursday introduces the programme name, the number of positions eliminated, the geographic scope, and the executive names of everyone being promoted into new roles. A caption programme that relies on a stable glossary will fail every time the strategic situation changes, which is every quarter at minimum.

Publication timelines

Training video is typically published to the LMS on a schedule that allows time for caption review. Earnings call recordings are published the same day as the live event. Investor day recordings may be published the same day or within 24 hours. M&A announcement videos are published the moment the announcement is made — often pre-market, before 8am Eastern, before the caption team has started work. The publication timeline for executive communications video is not governed by the training team's production schedule. It is governed by investor relations schedules, securities law disclosure timing, and the market's expectation that recordings will be available quickly. Caption review must happen inside that timeline, not after it.

Audience and legal exposure

Training video is viewed by employees. Executive communications video — specifically the public investor-facing category — is viewed by analysts, institutional investors, retail shareholders, journalists, and regulators. An inaccurate caption track on an earnings call recording is not just a compliance problem for employees with hearing disabilities. It is a potential accuracy problem for a hearing-impaired analyst or investor relying on the caption track to understand the company's guidance, its non-GAAP metrics, or its forward-looking statements. Securities law creates a separate set of reasons to care about transcript accuracy; the WCAG 2.1 AA accuracy standard creates an additional reason. The combination means that the acceptable error rate for earnings call captions is lower than for most other content types, and the consequences of getting it wrong are more visible.

Two compliance frameworks simultaneously

For most enterprise video, the compliance framework is a single question: is this employer-to-employee content (Title I) or public-entity-to-public content (Title II)? Executive communications splits this cleanly in most cases — but the split is not always obvious. An all-hands recording posted to the company intranet is Title I. The same recording published to the corporate website's News section is Title III. An investor day recording live-streamed on YouTube and subsequently hosted on the IR website is Title III for the public investor audience and, if employees are expected to watch it, Title I for the employee audience. The CEO's earnings commentary is Title III when it is on the IR website and Title I when the same content is included in the "recap for employees" internal communication. The compliance programme for executive communications cannot assume a single framework. It must classify each content destination separately.

Internal executive communications video categories

Internal executive communications video — content produced for an employee audience and distributed through internal channels — sits under ADA Title I, the same framework as training video. The employee communications captioning guide covers the general ownership and programme design questions for this category. This section focuses on what is specifically different about executive content within the internal communications bucket.

All-hands recordings: the exec comms layer

All-hands meetings and town halls are covered broadly in the employee communications guide. The executive communications-specific issues concentrate in the Q&A portion and the vocabulary change cycle. The prepared remarks in a quarterly all-hands meeting typically use vocabulary that was known at the time the executive team prepared their talking points. The live Q&A — unscripted, responding to employee questions — uses whatever vocabulary the executive reaches for in the moment. This is where product codenames, unreleased initiative names, and informal strategic vocabulary appear without having been entered into any glossary. A well-captioned prepared remarks section followed by a poorly captioned Q&A is the most common pattern in all-hands recordings at companies with partial caption programmes. The fix is a post-event glossary review: after every all-hands, the caption team reviews the Q&A transcript for new terms that appeared and adds them to the glossary before the next quarter's recording.

CEO strategic announcements

CEO video messages for strategic announcements — M&A, restructuring, major product pivots, leadership changes — are a distinct content type. They are typically short (3–7 minutes), high-production-value, and distributed quickly. The production chain is usually: CEO records with communications team present → communications team handles distribution through internal channels → IT or an AV production team handles the hosting platform. No one in this chain is typically thinking about captions until after the video is live. The accessibility coordinator, if one exists, learns about the recording from the intranet post, not from the production team. By the time the caption request reaches the caption team, the video has been live for hours or days. The fix is a production-chain integration: the corporate communications team must have a standing protocol to submit caption requests at the same time they schedule the recording session, not after it is published.

CFO and finance communications

CFO video communications — budget cycle updates, financial results explanations, cost management programme announcements — use a vocabulary register that is distinct from both the CEO strategic register and the general employee communications register. Budget cycle vocabulary (headcount envelope, operating budget, capital allocation, cost-centre reforecast, variance analysis, full-year guidance) differs from strategic vocabulary. Financial results explanations use the same GAAP/non-GAAP framework as earnings calls but in an internal context. Cost management announcements use the same restructuring vocabulary as M&A announcements but often in a more operational register (span of control, layers reduction, shared services consolidation, overhead reduction). Each of these registers requires glossary entries. A CFO who says "we are targeting 150 basis points of EBITDA margin expansion in the second half" has used four terms that general-purpose ASR will handle inconsistently — "basis points" may appear as "based points" or "base points," "EBITDA" may appear as "Ebita" or "e bit da," and "margin expansion" may be transcribed correctly but juxtaposed with the mangled metric name in a way that makes the sentence semantically wrong.

Board-level content distributed to employees

Board meeting recordings and board committee summaries are occasionally distributed to employees — typically the board chair's remarks or the summary of a board decision on compensation or strategy. This content carries ADA Title I obligations if it is distributed to employees as employer communication. Board vocabulary includes corporate governance terminology (fiduciary duty, duty of care, business judgment rule, audit committee, compensation committee, say-on-pay, proxy advisory firm recommendations) that is uncommon in any other internal content. Board member names and director backgrounds may appear in this content in ways that require specific pronunciation guidance. The distribution volume is low, but the legal sensitivity of the content — board deliberations and governance decisions — means that caption errors in board-distributed content have an elevated risk profile.

Earnings calls: the public webcast captioning problem

A quarterly earnings call is, by design, a public disclosure event. Under SEC Regulation FD (Regulation Fair Disclosure), material information about a public company must be disclosed simultaneously to all investors, not selectively to institutional investors. The earnings call live webcast and its recording are the primary mechanism for satisfying Reg FD. This means the earnings call recording is not internal communications video. It is a public webcast, hosted on the investor relations website, accessible to any member of the public — including members of the public with hearing disabilities who rely on captions to access audio-only or audio-primary content. The earnings call recording is ADA Title III content. The WCAG 2.1 AA standard applies. The DOJ web accessibility final rule under Title III (March 2024, effective June 2024 for large businesses) covers websites of places of public accommodation, and the investor relations page of a public company's website is within scope.

What an earnings call contains

A standard quarterly earnings call has four segments, each with distinct vocabulary requirements. Understanding the vocabulary breakdown by segment is necessary for building an adequate glossary and understanding where auto-caption failure rates are highest.

The safe harbor statement: Legal counsel reads a prepared safe harbor disclaimer identifying the forward-looking statements in the prepared remarks, listing specific risk factors, and directing listeners to the company's most recent 10-K and 10-Q filings. The safe harbor statement is read at high speed from a prepared legal text. It uses securities law vocabulary (forward-looking statements, material risks, GAAP, non-GAAP, Regulation G, 8-K, 10-K, 10-Q, risk factors, results of operations, liquidity and capital resources, actual results may differ materially) at a rate and rhythm that general-purpose ASR cannot handle accurately. The safe harbor statement is also the statement that a plaintiff's attorney would review first if there were ever a securities fraud claim involving the transcript. An inaccurate caption track of the safe harbor statement — one that drops the forward-looking statement identification or mangles the risk factor language — is not a trivial QA miss.

The CEO prepared remarks: The CEO discusses business performance, strategic priorities, and qualitative commentary on the quarter. Vocabulary: product name mentions, initiative names, go-to-market programme names, competitive framing (naming competitors who may have unusual proper nouns), customer names (reference accounts mentioned by name), and strategic framework vocabulary (the specific language the company uses to describe its market position and growth model). This section is typically more legible to ASR than the safe harbor statement, but proper noun accuracy depends heavily on whether the company's specific product names, customer names, and initiative names are in the glossary.

The CFO prepared remarks: The CFO presents the financial results in detail, moving line-by-line through the income statement, balance sheet, and cash flow statement, then providing guidance for the next quarter and the full year. This section has the highest concentration of financial vocabulary and the highest auto-caption failure rate. See the vocabulary breakdown below.

The analyst Q&A: Sell-side analysts from major investment banks ask questions. Analyst names (Raymond James, Jefferies, Piper Sandler, Baird, William Blair — the analyst names, not just the firm names) will appear in speaker identification sections. The vocabulary in Q&A is unscripted, includes specific financial model terminology that analysts bring to their questions (SaaS metrics, cohort analysis vocabulary, unit economics framing), and includes the CEO and CFO responding to specific competitive or operational questions with vocabulary that was not in the prepared remarks.

Financial vocabulary that auto-captions mangle

The following vocabulary categories represent the highest failure-rate domains for general-purpose ASR on earnings call content. Each item includes the correct form and the common auto-caption error.

GAAP/non-GAAP metrics: "Non-GAAP" is the single most commonly mangled term on earnings calls. Auto-captions produce "non-gap," "non-gab," "non-GAAP" (sometimes correctly), and occasionally "non-cap." "GAAP operating income" may be correctly transcribed while "non-GAAP operating income" in the same sentence is not. "Adjusted EBITDA" may appear as "adjusted Ebita," "adjusted E B I T D A" (spelled out), "adjusted a bit ah," or occasionally correct. "EBITDA margin" has similar failure modes. The common pattern is that the acronym is transcribed differently each time it appears in the recording — sometimes correctly, sometimes not — which produces a transcript that appears to use multiple different terms where the speaker used one.

Financial statement line items: "CapEx" appears as "cap X," "cap ex," "cape X," or correct. "OpEx" appears as "op X," "op ex," "oh pecs," or correct. "Gross margin" is usually transcribed correctly, but "gross margin expansion" is often transcribed as "gross margin X pension." "Free cash flow" is often transcribed correctly, but "free cash flow conversion" may be mangled. "Working capital" is usually correct. "Days sales outstanding" may appear as "D.S.O." or "days sales outstanding" depending on whether the speaker abbreviates. "Accounts receivable" is usually correct. "Deferred revenue" is usually correct. "Remaining performance obligations" (RPO) almost always appears as "R.P.O." or "Arco" rather than spelled out, depending on the speaker.

SaaS and recurring revenue metrics: "Annual recurring revenue" (ARR) appears as "A.R.R.," "are," or correct. "Net revenue retention" (NRR) appears as "N.R.R.," "narrow," "near," or correct. "Gross revenue retention" (GRR) appears as "G.R.R.," "grrr," or correct. "Customer acquisition cost" (CAC) appears as "C.A.C.," "cack," "stack," or correct. "Net dollar retention" (NDR) — sometimes used interchangeably with NRR by different companies — has the same issue as NRR plus the additional confusion of being a near-homophone for "N.D.R." "Lifetime value" (LTV) appears as "L.T.V.," "love," "Ltv," or correct. "Rule of 40" is usually correctly transcribed. "Magic number" (sales efficiency metric) is usually correct but contextually confusing if not captioned in financial context.

Guidance and consensus language: "Consensus" is usually correctly transcribed. "Beat consensus" is usually correct. "Miss consensus" is usually correct. "Guidance range" is usually correct. "We are guiding to revenue of 185 to 190 million dollars" is usually transcribed correctly when the numbers are well-articulated, but may produce "185 to 190" without the "million dollars" if the speaker trails off. "At the midpoint of our guidance range" is usually transcribed correctly. "Our full-year revenue guidance of 745 to 755 million dollars" is usually transcribed correctly for the numbers but "full-year revenue guidance" may be mangled to "full year revenue guidance" or "full your revenue guidance." The distinction between "our guidance is above consensus" and "our guidance is above the consensus range" matters semantically and is often incorrectly transcribed when spoken quickly.

Shareholder and securities vocabulary: "Earnings per share" (EPS) appears as "E.P.S.," "EPA," "ESP," "eps," or correct. "Diluted EPS" may appear as "diluted E.P.S." or "diluted eps" or "diluted ease peas." "Basic shares outstanding" is usually correct. "Weighted average diluted shares" may be mangled at any word. "Share repurchase" is usually correct. "Buyback" is usually correct. "Dividend per share" is usually correct. "Yield" is usually correct in context. "Total shareholder return" (TSR) appears as "T.S.R." or "teaser" or correct.

Regulatory and disclosure vocabulary: "Regulation FD" appears as "Regulation F.D.," "Regulation FD," or occasionally "regulation food" (phonetic). "Material non-public information" is usually correct when spoken slowly, but "MNPI" appears as "M.N.P.I.," "mini," "Minnie," or correct. "8-K" appears as "eight K," "8 K," "A Kay," or correct. "10-K" appears as "ten K," "10 K," "ten kay," or correct. "10-Q" appears as "ten Q," "10 Q," "ten cue," or correct. "Safe harbor" is usually correct.

Analyst and bank names: Analyst names are a particular failure point because they are proper nouns that appear in "the next question comes from [Name] at [Firm]" transitions. "Piper Sandler" may appear as "Piper Sandler" (correct), "Piper Sander," "Piper sandler," or "Piper send her." "Jefferies" may appear as "Jeff rees," "Jeff ries," or correct. "Baird" may appear as "bared" or "bard" or correct. "William Blair" may appear as "William Blair" (correct) or "William Blaire." "Raymond James" is usually correct. "Deutsche Bank" may appear as "Deutsche Bank" (correct) or "Deutch Bank" or "Dutch Bank." These names appear in the transition segments between each analyst question, so an uncorrected analyst name produces an incorrect speaker identification label for every question from that analyst.

The live webcast versus the recorded archive

Earnings calls are typically conducted as live webcasts — the management team is on a phone call with a conferencing bridge that is simultaneously broadcast on the internet via a webcast platform. Listeners can attend the live event or access the recorded archive afterward. The live event and the recorded archive have different caption characteristics.

During the live event, the webcast platform may display real-time captions generated by the conferencing platform or by a professional CART (Communication Access Realtime Translation) captioner. If the company uses Zoom Webinars, Microsoft Teams Live Events, Webex Webinars, or a dedicated earnings call platform (Q4 Inc., Notified by West), the live caption capability varies by platform. Professional CART captioning is a separate engagement with a human captioner who transcribes in real time — this is the only live captioning method that achieves WCAG SC 1.2.1-equivalent accuracy for live content. Live platform auto-captions (Zoom, Teams) achieve 80–88% accuracy on clear executive speech and substantially lower accuracy on financial vocabulary. Neither live captioning method produces a caption track that can be saved and attached to the recorded archive — they are displayed during the live event and then discarded.

The recorded archive, posted to the IR website after the call, requires a separate caption track — a correctly formatted SRT or VTT file attached to the video player. This caption track must meet WCAG SC 1.2.2 (99% accuracy for pre-recorded content). The live captioning, whether CART or auto-captions, does not satisfy this requirement for the recording. The recording requires its own caption workflow.

Investor day presentations

An investor day (also called a capital markets day or analyst day) is a multi-hour event in which the management team presents the company's strategy, product roadmap, financial model, and long-term targets to investors and analysts. Investor day recordings are public webcasts — they are streamed live on the IR website and YouTube, and the recordings are archived on the IR website alongside the slide deck and the prepared transcript. They are ADA Title III content, and they present the most demanding captioning challenge in the executive communications category.

Duration and multi-speaker complexity

Investor days typically run three to eight hours. A typical format: a welcome from the CFO or Head of Investor Relations (15 minutes), a CEO keynote on strategy (45–60 minutes), individual product area or geographic segment deep-dives from business unit leaders (30–45 minutes each), a financial model presentation from the CFO (45–60 minutes), and a closing panel Q&A with multiple executives responding to analyst questions simultaneously. A 6-hour investor day recording contains approximately 400,000 to 500,000 words of spoken content. Captioning 400,000 words at WCAG quality requires either a professional captioning service or a high-accuracy ASR pipeline with a comprehensive production glossary and post-processing review. It cannot be done with raw auto-captions.

Speaker identification is a compounding problem at investor days. Five to fifteen executives present, plus a moderator, plus 20–40 analysts in the Q&A. Speaker labels must be accurate for the caption track to be useful. If the speaker identification labels are wrong — if the CEO's remarks are labeled as the CFO's, or an analyst from Baird is labeled as an analyst from Goldman Sachs — the caption track is misleading even if the words themselves are correctly transcribed. Speaker identification requires either a manual review pass or a speaker diarization system trained on the speakers' voices.

Investor day vocabulary categories

Market sizing vocabulary: TAM (total addressable market), SAM (serviceable addressable market), SOM (serviceable obtainable market). "TAM" appears as "tam," "T.A.M.," or correct. "SAM" appears as "Sam," "S.A.M.," or correct. "SOM" appears as "sum," "some," "S.O.M.," or correct. The phrase "our total addressable market of 45 billion dollars is underpenetrated at current attach rates" is typically transcribed correctly for the words but the SAM/SOM terms within it may be mangled.

SaaS business model vocabulary: Net Revenue Retention (NRR) — addressed above under earnings calls. Gross Revenue Retention (GRR) — same issues. Dollar-based net revenue expansion (DBNRE) — appears as "D.B.N.R.E.," "deal B N R E," or occasionally correct. Annual Contract Value (ACV) — appears as "A.C.V.," "acetyl," or correct. Average Revenue Per User (ARPU) — appears as "ARPU," "ar poo," or correct. Logo retention — usually correct. Upsell/cross-sell motion — usually correct. Land and expand — usually correct. "Net dollar retention of 120 percent" is usually correctly transcribed but "our trailing twelve-month NRR is 118 to 122 percent" may mangle "NRR" in any of the ways described above.

Go-to-market vocabulary: GTM (go-to-market) appears as "G.T.M.," "Jim," "goto market," or correct. "Product-led growth" (PLG) appears as "P.L.G.," "plug," "pull G," or correct. "Sales-led growth" is usually correct. "Channel partner" is usually correct. "Reseller motion" is usually correct. "Direct sales" is usually correct. "Enterprise motion" is usually correct. "SMB motion" appears as "S.M.B. motion," "small and medium business motion," or correct. "Mid-market motion" is usually correct. "Self-serve" is usually correct. "Hybrid motion" (PLG plus sales-led) is usually correct.

Financial model vocabulary: "Rule of 40" (revenue growth rate plus free cash flow margin) is usually correctly transcribed. "FCF margin" appears as "F.C.F. margin," "FCF margin" (correct), or "efsy margin." "Operating leverage" is usually correct. "Gross margin expansion" — addressed above. "Sales and marketing efficiency" is usually correct. "Research and development intensity" is usually correct. "General and administrative leverage" — "G and A leverage" may appear as "general and administrative leverage" (correct) or abbreviated inconsistently. "Long-range plan" (LRP) appears as "L.R.P.," "Larp," or correct. "Three-year CAGR" appears as "three-year C.A.G.R.," "three-year cager," "three-year kayjer," or correct.

Product and technology vocabulary: Product names from the company's product suite, API vocabulary (integration depth, surface area, webhook, connector), platform architecture vocabulary (multi-tenant, single-tenant, cloud-native, on-premise deployment), and competitive differentiation vocabulary vary by company but collectively represent a vocabulary set that is guaranteed to fail ASR without a product-specific glossary. Investor day is also where the company's roadmap products are mentioned by name for the first time — products that are not yet in any glossary, not yet publicly documented, and often under NDA until the investor day itself. A roadmap product revealed at investor day has exactly zero coverage in any existing glossary.

CEO strategic communications vocabulary

CEO strategic communications — M&A announcements, restructuring communications, strategic pivot announcements — represent the highest-volatility vocabulary category in executive communications. Each event introduces a vocabulary set that is unique to that transaction, that programme, or that initiative, and that vocabulary is unknown to any captioning system before the announcement is made. The vocabulary is also high-consequence: the precise language of a merger announcement, a reduction in force, or a strategic pivot carries specific legal and relational meaning for the employees, investors, and markets receiving the communication.

M&A announcement vocabulary

When a company announces an acquisition or merger, the CEO video message uses a vocabulary set that is specific to the transaction. The target company name is the first vocabulary problem: if the company being acquired has a name that is unusual (a proper noun from a non-English language, a portmanteau, a stylised spelling, a number or symbol in the name), ASR will produce an incorrect transcription. The phrase "we are pleased to announce that we have entered into a definitive agreement to acquire Acme Systems" requires the ASR system to correctly transcribe "definitive agreement" (usually correct) and "Acme Systems" (only correct if "Acme Systems" is in the glossary — "Acme" is a common word but "Acme Systems" as a proper noun entity may be transcribed as "Acme systems" with lowercase, "Acme sis tems," or occasionally something more divergent depending on the target name).

Transaction structure vocabulary: "Cash and stock consideration" is usually correctly transcribed. "Merger consideration of 42 dollars per share" is usually correctly transcribed for the number but "merger consideration" may appear as "merger consideration" (correct) or "merger con side ration." "The acquisition is expected to close in the second half of 2026 subject to customary closing conditions, including regulatory approval" uses vocabulary that is usually transcribed correctly individually but may be garbled in sequence at reading speed. "HSR filing" appears as "H.S.R. filing," "easier filing," "hazer filing," or correct. "Hart-Scott-Rodino" is almost never transcribed correctly by ASR — it appears as "Hart Scott Rodino," "Heart Scott Rodino," or any number of phonetic approximations, none of which are the correct filing name. "Second request" (HSR second request, the additional information request from the reviewing agency) appears as "second request" (usually correct, but the "HSR" context may be garbled).

Financial impact vocabulary: "Accretive to non-GAAP EPS" is a phrase that appears in virtually every M&A announcement and fails in multiple ways: "accretive" appears as "a creative," "acretive," "a creditive," or correct; "non-GAAP" has the issues described above; "EPS" has the issues described above. "Dilutive to GAAP EPS in the near term" has the same problem plus "dilutive" appearing as "dilute ive" or "diluitive." "Synergies" is usually correctly transcribed. "Revenue synergies" is usually correct. "Cost synergies" is usually correct. "Integration costs" is usually correct. "Amortization of acquired intangibles" — "amortization" is usually correct, "intangibles" is usually correct, but the full phrase at reading speed may be garbled.

Integration vocabulary: The integration plan section of an M&A communication uses an integration team name (often created for the announcement, e.g., "Project Horizon Integration Team") that is not in any glossary. The integration workstream names (finance integration, technology integration, go-to-market integration, people integration) are usually correctly transcribed. The integration leader names are new vocabulary that requires glossary entries.

Restructuring and reduction in force vocabulary

Restructuring communications are particularly high-consequence for caption accuracy because they describe employment changes that directly affect the people watching the video. A CEO describing a reduction in force who uses vocabulary that is garbled in the caption track creates an accessibility failure for hearing-impaired employees who are relying on captions to understand whether their job is affected.

RIF vocabulary: "Reduction in force" (RIF) appears as "R.I.F.," "riff," "reef," or correct. "Headcount reduction" is usually correctly transcribed. "Position elimination" is usually correct. "Impacted employees" is usually correct. "Transition support" is usually correct. "Severance package" is usually correct. "Outplacement services" — "outplacement" appears as "out placement" (usually correct) or "out-placement." "Separation agreement" is usually correct. "WARN Act" (the Worker Adjustment and Retraining Notification Act, which requires 60-day notice for large-scale layoffs at qualifying employers) appears as "warn act," "WARN act," "worn act," or correct. "60-day WARN notice" — "WARN" in this context is capitalized in the statute and often mispronounced and misidentified by ASR as the ordinary English word "warn." The WARN Act is a specific statutory obligation that many affected employees will be researching; an ASR failure on "WARN Act" that produces "warn act" without the proper noun identification creates unnecessary confusion.

Programme naming: Restructuring programmes are often given internal programme names ("Project Simplify," "Operational Excellence Initiative," "Efficiency Programme," "Transformation 2026") that appear in the announcement video and in subsequent communications. These names are not in any glossary before the announcement. If the programme name is announced in a CEO video, it needs to be added to the glossary immediately and applied to the caption track before the video is published.

Strategic pivot and product evolution vocabulary

Strategic pivot communications — announcements of new business models, market exits, new market entries, product line consolidations, or platform transitions — use vocabulary that is specific to the new direction and that may not align with any existing glossary. A company pivoting from a horizontal tool to a vertical platform uses "verticalization" vocabulary. A company shifting from sales-led to product-led growth uses PLG vocabulary. A company exiting a product category uses "sunset," "end-of-life" (EoL), "migration path," "transition timeline," and "replacement product" vocabulary. Each of these pivots introduces a vocabulary set that is new relative to the company's existing glossary and that requires explicit glossary entries before the announcement video is captioned.

The 48-hour accuracy window

The phrase "48-hour accuracy window" describes a specific constraint in investor communications video: the time between the live earnings call and the publication of the archived recording on the IR website is typically two to six hours, and within that window the caption track must be reviewed and attached. Once the recording is live on the IR website without a caption track — or with an auto-generated caption track that has not been reviewed — any delay in correction is a period of non-compliance. For earnings calls, that window is smaller than for almost any other enterprise content type.

Why the window is short

Earnings call recordings are published quickly because investors and analysts want access to the replay. The standard expectation in investor relations is that the webcast recording will be available within one to two hours of the call ending. Some companies publish within 30 minutes. The investor relations team is managing multiple simultaneous deliverables in the post-call window: the 8-K filing with the press release, the transcript (often provided by a third-party transcription service), the slide deck archive, the webcast replay link, and the IR website update. Caption review is not a standard item in the typical IR team's post-call checklist. Adding it requires both making it a checklist item and ensuring that someone with the ability to review caption accuracy is available and alert during the post-call window — which may be 6am Eastern on a weekday, or mid-afternoon, depending on when the call is scheduled.

The transcript as a caption source

Many earnings call setups include a professional transcript produced by a third-party transcription service (Refinitiv, Factset, Bloomberg, Q4 Inc., or a dedicated earnings transcript service). These transcripts are produced by human transcribers who review the audio and produce a word-for-word written record of the call. They are accurate — professional earnings transcripts produced by the major services typically exceed 98% accuracy on financial vocabulary, with structured analyst identification and prepared remarks labeling. The question is whether this transcript can be used to produce a caption track for the webcast recording.

A flat transcript is not a caption file. A caption file requires accurate timestamps for each line of text — the SRT or VTT format requires a start time, end time, and caption text for each segment. A professional transcript without timestamps must be converted to a timed caption file. This conversion can be done by aligning the transcript text to the audio of the recording and generating timestamps — a process that takes 30–90 minutes for a 60-minute earnings call depending on the tool used and the alignment accuracy. If the IR team already commissions a professional transcript, the additional step of converting that transcript to a timed caption file for the webcast is a tractable extension of the existing transcript workflow. The key requirement is that the conversion happens inside the same post-call window as the transcript itself, not days later.

The caption review checkpoint

Whether the caption source is a professional transcript converted to SRT/VTT or an ASR-generated draft reviewed by a human, the caption review checkpoint must be scheduled in the IR team's post-call runbook. A runbook item with a named owner, a time budget (typically 30–60 minutes for a 60-minute call review), a review platform (the video hosting platform's caption upload interface), and a confirmation step that the caption track is live before the recording is marked as published is the operational implementation of the accuracy window requirement. Teams that do not have a formal earnings call runbook — who treat the post-call period as an informal sprint to get things live — will find it difficult to add caption review as a step without first formalising the other post-call deliverables.

Retroactive correction and the compliance window

If the earnings call recording goes live without a caption track, or with an uncorrected auto-generated track, the first question the IR team faces is: how quickly can we get a corrected caption track live? For WCAG SC 1.2.2 compliance, the pre-recorded multimedia content must have synchronised captions. "Pre-recorded" means content that is not live; the archived recording is pre-recorded and must have captions. A recording that is live without captions is non-compliant for the duration that it remains captionless. A corrected caption track uploaded the next day restores compliance from the time of upload, but does not retroactively repair the compliance gap from the previous day. This is not a theoretical risk: investors and analysts who access the recording in the first hours after publication include the people with the most direct interest in the content — the analysts who cover the stock and the institutional investors who may be making trading decisions based on the call. If any of those individuals have a hearing disability and rely on captions, they are the people most likely to access the recording immediately after publication and to find it inaccessible.

Platform workflows for executive communications

Earnings call platforms

Earnings calls are conducted on one of several platform types: dedicated investor relations webcast platforms, general enterprise conferencing platforms, or a hybrid of both.

Q4 Inc. (formerly IR Webcasting): Q4 Inc. is the dominant dedicated investor relations webcast platform. Q4 handles the live call bridge, the live webcast stream, the webcast recording, and the IR website hosting of the archived recording. Caption workflow: Q4 does not auto-generate caption tracks for archived recordings by default. The caption file (SRT or VTT) must be uploaded to the Q4 platform after the call. Q4's platform supports caption file upload via the IR website management console. The caption file must match the recording's timing — it cannot be a flat transcript without timestamps. Q4 can accept VTT or SRT format. The upload is the IR team's responsibility; Q4 does not provide captioning services as a standard platform feature.

Notified by West (formerly NASDAQ Corporate Solutions Webcast): Notified is the second major dedicated earnings webcast platform. Caption support is similar to Q4: the platform hosts the archived recording and supports caption file upload. Notified has offered real-time captioning as an optional add-on for live events; the archived recording caption track is a separate deliverable. Caption file format requirements should be confirmed with the Notified account representative at the start of each quarter.

Zoom Webinars: Some companies conduct earnings calls via Zoom Webinars, particularly post-pandemic. Zoom Webinars generates real-time auto-captions during the live event (Zoom AI Companion, formerly Otter.ai-based) and produces a live transcript. The live transcript is not a caption track for the recording — it is a separately generated document with different timestamp alignment. Zoom Cloud Recordings (the recorded version) support caption file upload in the video settings. The Zoom-generated auto-caption file (available in TXT or VTT format from the recording) has the financial vocabulary accuracy problems described above and requires review before use. The Zoom Webinar recording can be published directly from Zoom to an external link; if the external link is the IR website embed, the caption configuration in the Zoom recording settings governs what the viewer sees.

Webex Webinars: Webex Webinars (Cisco Webex Events) is used by some enterprise companies for earnings calls. Webex provides real-time closed captioning powered by Cisco's Voicea/WebexAI transcription engine. The live caption accuracy on financial vocabulary is comparable to Zoom — adequate for general speech, poor on acronyms and financial metrics. The Webex recording can be hosted on the Webex platform or downloaded and uploaded to the IR website video host. Caption files from Webex recordings are available in VTT format from the recording management console. The same accuracy issues apply; the file should be reviewed before attachment.

IR website video hosting

The investor relations website typically hosts archived earnings call recordings and investor day recordings using one of several video hosting platforms. The caption configuration lives at the hosting layer, not the IR website CMS layer.

Wistia: Wistia is widely used for IR website video hosting due to its customisable player, chapter marker support, and clean embedding. Wistia supports SRT and VTT caption upload directly in the video management console (Media > Captions). Multiple language caption files can be uploaded. The caption track is delivered through the Wistia player; if a video is embedded in the IR website via Wistia embed code, the caption track configured in Wistia is what the viewer sees. Caption tracks must be manually uploaded — Wistia does not auto-generate caption files, though Wistia AI (Wistia's transcription feature) can produce a draft transcript that requires review. The accuracy of Wistia AI transcription on financial vocabulary is comparable to other general-purpose ASR — adequate for baseline review, inadequate as a production caption track without correction.

Vimeo: Vimeo is used for some IR website video hosting. Vimeo supports SRT and VTT caption upload in the Video Settings > Distribution > Subtitles & Captions section. Vimeo also has an auto-caption feature (Vimeo AI captioning) that produces a draft transcript for review. Same accuracy caveats as Wistia AI. The uploaded caption file is delivered through the Vimeo player.

YouTube (unlisted): Some companies host earnings call recordings on YouTube as unlisted videos embedded in the IR website. YouTube's auto-generated captions are among the most commonly used and least reviewed caption tracks for earnings content — the video is on YouTube, YouTube generates captions automatically, and the IR team assumes the captions are covered. YouTube auto-captions on earnings call content will have all the financial vocabulary failure modes described above. YouTube's auto-caption quality on financial terminology is particularly inconsistent because the YouTube ASR model is optimised for the general YouTube content mix, not for financial analyst calls. Correcting YouTube captions requires accessing YouTube Studio, editing the auto-generated caption file, and saving the corrected version. If the earnings recording is published on YouTube and the IR team does not own the YouTube account (common when the marketing or communications team owns the company YouTube channel), caption correction may require a cross-team workflow.

The Q4 or Notified hosted player: If the archived recording is hosted directly on the Q4 or Notified platform rather than on a third-party video host, the caption file is managed within that platform's console. The workflow is platform-specific and should be confirmed with the platform's technical support at the start of the quarter.

All-hands and CEO message platforms

Internal executive communications video is typically hosted on the same platforms as general employee communications video. The platform workflows for Zoom, Teams, Vimeo, and Wistia for internal use are covered in the employee communications captioning guide. The executive-specific additions to platform workflow are:

The CEO message recording session: CEOs typically record video messages using a combination of production tools: an in-office setup with a dedicated camera and teleprompter, a remote recording session via Zoom or Teams (recorded locally and edited), or a professional video production team for high-production announcements. The caption workflow depends on which recording method was used. Teleprompter-based recordings using a prepared script can produce a caption file from the script before the recording is even edited — the script is the source text, and timestamps can be added in post-production. Remote recording sessions (Zoom/Teams) produce a recording file that goes through the same ASR captioning workflow as any recorded meeting. Professional production sessions should include caption production in the production scope from the beginning.

Board and leadership distribution channels: Executive content distributed to the board or to a restricted leadership group may use a different hosting environment than the general company intranet — a board portal (Diligent, BoardVantage, Nasdaq Boardvision) or a secure document sharing platform. These platforms may or may not support embedded video playback with caption tracks. If board portal video hosting does not support caption file upload, the caption track must be delivered as a separate file (e.g., a plain text SRT attached to the board package) with instructions for use. This is a degraded but defensible accommodation for a small audience in a secure environment.

Glossary architecture for executive content

The glossary architecture guide covers the structural design of a production caption glossary: term types, confidence scoring, phonetic pairs, and the relationship between the glossary and the ASR pipeline. This section covers the specific architecture requirements for executive communications content, which differ from the L&D training glossary in four ways: volatility, sensitivity, multi-register structure, and update ownership.

The four vocabulary categories for executive content

Category 1 — Stable financial vocabulary: Financial statement terminology, securities law vocabulary, and earnings call structure vocabulary that remains constant across quarters. This is the category most analogous to the L&D training glossary — terms that can be entered once and maintained as needed. Examples: EBITDA, GAAP, non-GAAP, EPS, guidance range, Regulation FD, 10-K, 10-Q, 8-K, safe harbor. These terms should be built into the executive communications glossary at programme inception and reviewed annually. This category has the highest confidence scores because the pronunciation is fixed and the context is always the same.

Category 2 — Stable company and product vocabulary: Company name, product names, brand names, executive names, and institutional investor names that appear consistently across earnings calls and investor days. This category overlaps with the L&D training glossary for company and product names, but extends to include sell-side analyst names, institutional investor names (Vanguard, BlackRock, Fidelity, T. Rowe Price — names of major shareholders whose representatives ask questions on earnings calls), and financial media outlet names. Executive names require phonetic entries for any name with non-obvious pronunciation. Analyst names should be entered for the top 10–15 analysts who regularly attend the company's earnings calls — a list that can be obtained from the investor relations team's regular attendee tracking.

Category 3 — Quarterly-volatile vocabulary: Terms that are specific to the current quarter's strategic situation, financial results, and initiatives. This category requires a new vocabulary pass before every earnings call. Items in this category include: the specific revenue guidance figures for the coming quarter (the numbers themselves, which the ASR will usually transcribe correctly, but also the initiative names associated with the guidance), new product or feature names mentioned for the first time, new strategic programme names, new customer reference accounts mentioned by name, and new competitive comparisons (if a competitor is named for the first time on this quarter's call). Quarterly-volatile vocabulary entries have a shorter useful life — they may be accurate for one earnings call and irrelevant by the next. The glossary management system must support marking terms with a "last verified" date and a "context" field indicating which quarter the term was relevant to.

Category 4 — Event-specific vocabulary: Terms that are specific to a single event and are unlikely to recur. M&A target company names, restructuring programme names, strategic initiative names announced at investor day, acquisition integration team names, and board-approved executive names for new leadership announcements fall into this category. These terms should be entered in the glossary immediately when the event is scheduled or announced, and should be marked as "event-specific" with the event date and name. After the event, these terms may be promoted to Category 2 (if the target company name becomes the acquiree's brand in ongoing use) or archived (if the programme name is used only in the announcement context).

The confidentiality constraint on pre-announcement glossary updates

M&A transactions, major restructuring programmes, and strategic pivots are typically under strict confidentiality before the public announcement. The vocabulary specific to the transaction — the target company name, the programme name, the deal structure terms — cannot be entered into a shared glossary before the announcement date without creating a confidentiality risk. If the caption glossary is accessible to the caption vendor, adding the target company name before the announcement is effectively disclosing material non-public information to the vendor. This constraint has a practical resolution: a restricted glossary that is held within the investor relations and legal teams, not shared with the caption vendor, and merged into the production glossary at the moment of the announcement. Some caption systems support user-managed glossary entries that can be used for a specific job submission without being entered into the shared account glossary — this architecture allows M&A vocabulary to be applied to the announcement recording without exposing it to the vendor's glossary system.

Glossary ownership for executive content

The caption programme governance policy assigns glossary ownership by content category. For executive communications content, the ownership structure differs from the L&D training glossary. The recommended ownership assignment is: Category 1 (stable financial vocabulary) — caption programme manager or accessibility coordinator, reviewed annually with input from IR or Finance. Category 2 (stable company and product vocabulary) — caption programme manager, with updates triggered by product naming changes, executive changes, or major investor announcements. Category 3 (quarterly-volatile vocabulary) — investor relations team, with a quarterly update workflow timed to earnings call preparation. Category 4 (event-specific vocabulary) — the team that owns the event (IR for M&A, Corporate Communications for restructuring, strategy team for investor day), with entries submitted to the caption programme manager for glossary entry at the announcement date.

The quarterly update workflow for Category 3 vocabulary should be triggered when the company's IR team begins preparing earnings call talking points — typically two to four weeks before the scheduled earnings call. A vocabulary submission form or shared document in which the IR or Finance team enters new terms, guidance figures with their specific phrasing, and new initiative names is the most practical implementation. The caption programme manager reviews the submissions, formats them as glossary entries, and uploads them to the ASR system before the earnings call recording is submitted for captioning.

Disability accommodation requests for earnings content

A disability accommodation request for earnings call content does not come to the L&D team or the accessibility coordinator through the normal HR accommodation channel. It comes to the investor relations team through an investor inquiry channel — an IR contact email, an IR hotline, or a direct message to the Head of Investor Relations. The IR team is not typically equipped to respond to accessibility accommodation requests. They are equipped to respond to investor questions about financial results, guidance, and investor events. This mismatch creates a response failure unless the programme design includes a specific protocol for routing accessibility requests from IR to the caption programme.

Who makes these requests and what they need

Hearing-impaired investors and analysts: A buy-side analyst or retail investor with a hearing disability who relies on captions to access the earnings call recording may contact the IR team if the caption track is missing or inaccurate. The request is typically for a corrected caption file for the archived recording, or for a captioned version of the live webcast for future events. The response from the IR team should be: acknowledgment within one business day, escalation to the caption programme manager or accessibility coordinator, and delivery of a corrected caption track within the timeframe consistent with the company's ADA accommodation response practices (typically within a reasonable time, often 3–5 business days for a corrected file). For future events, the response is: confirmation that future recordings will include a caption track that meets WCAG SC 1.2.2, and for live events, an option for real-time CART captioning if the company's live webcast setup supports it.

Hearing-impaired employees: Employees who are expected to watch the earnings call or investor day recording as part of their job — finance team members, investor relations staff, business partners briefed on strategy — may make an accommodation request through HR. This request may come through the standard ADA Title I accommodation process (request to HR, review by HR and manager, accommodation plan). The accommodation for this request is a compliant caption track on the recording. If the recording already has a compliant caption track, the request is already satisfied. If the recording has no caption track or an inaccurate auto-generated track, the accommodation requires a corrected file. The employee accommodation channel (HR) and the investor accommodation channel (IR) may each receive requests about the same recording; the programme design must ensure both channels route to the same resolution process.

Hearing-impaired board members: A board director with a hearing disability who requires captions for board meeting recordings is entitled to effective access to board materials as part of their role as a director. Board member accommodation is not an ADA Title I matter in the same way as employee accommodation — board members are not employees in most jurisdictions — but the company has an obligation to ensure its board members can participate effectively in governance. A board director's request for captioned board recordings should be routed through the Corporate Secretary's office and accommodated through the same caption production channel as other executive content.

The Title III response timeline

For investor-facing earnings call content (ADA Title III), the accommodation response is not governed by the ADA Title I accommodation process (which has specific HR-managed timelines and documentation requirements). Title III requires "equal enjoyment of goods, services, facilities, privileges, advantages, and accommodations." For a pre-recorded webcast, this means the caption track should be present when the content is published, not reactive to a request. If the caption track is missing at publication and an investor contacts the IR team to request it, the company is already in a position of non-compliance; the request is a notification of non-compliance rather than a prospective accommodation request. The response is to provide the corrected caption track as quickly as possible and to correct the root cause (the production workflow that allowed the recording to go live without a compliant caption track).

Eight failure modes in executive communications captioning

1. Treating the earnings call transcript as a substitute for a timed caption file

Professional earnings call transcripts — produced by Refinitiv, Factset, Bloomberg, or the webcast platform's transcript service — are accurate but are flat text documents without timestamps. A flat transcript cannot be used as a caption track because SRT and VTT formats require start and end timestamps for each caption segment. Teams that discover the earnings call recording has no caption track sometimes upload the plain-text transcript as a caption file in the wrong format, or insert it as a static overlay (a visible text box on the video), or link to it as a separate document adjacent to the video. None of these are equivalent to WCAG SC 1.2.2 synchronised captions. The fix is to convert the professional transcript to a timed caption file by aligning it to the audio — a step that takes 30–90 minutes and can be built into the post-call workflow.

2. Using live CART captions for the recording

Some companies commission a professional CART captioner for the live earnings call or all-hands event. CART (Communication Access Realtime Translation) is a human stenographic service that produces real-time captions during a live event at 99%+ accuracy. The CART output is displayed during the live event on a separate stream or web page. The common failure is assuming that the CART output can be saved and used as the caption track for the recording. CART files are typically delivered in a proprietary format (RTF or CRF from the CAT software the stenographer uses) with timestamps that reference real-world clock time rather than offsets from the recording start. Converting a CART file to a properly timed VTT or SRT for a recorded webcast requires realigning the timestamps to the recording's internal clock. If the recording and the CART session started at different times — which they often do, because the webcast platform starts recording when participants join and the CART session starts when the captioner connects — the alignment will be off. This is a solvable problem but requires explicit attention in the post-call workflow.

3. The auto-caption "we covered it" assumption for all-hands recordings

The most common failure mode in executive communications captioning is the identical pattern described in the employee communications guide: the team enables live Zoom or Teams auto-captions for the all-hands event, and then assumes the recording has captions because the live event did. Live platform auto-captions are not saved to the recording. The Zoom cloud recording of the all-hands does not include the live caption stream; it requires a separate caption file to be uploaded to the recording's caption settings. This failure mode affects every organisation that uses live auto-captions and has not explicitly built a post-event captioning step into the all-hands workflow. It is the single most common source of uncaptioned executive communications content.

4. The IR team publishes before caption review is complete

The post-earnings call rush to get the recording, transcript, press release, and 8-K live creates time pressure on every deliverable in the IR workflow. Caption review is the newest item in the workflow and has no historical precedent to give it priority. The result: everything else goes live first, and the caption upload is deferred to "later today" or "we'll get it up tomorrow." Once the recording is live without a caption track, the non-compliance clock starts. The fix is a workflow gate: the recording cannot be published until the caption track is attached. This gate requires IR team buy-in and a caption delivery that fits inside the same window as the other post-call deliverables.

5. Quarterly earnings vocabulary not updated before the call

The caption glossary is not updated before the earnings call because the IR team does not own the glossary and does not have a workflow prompt to update it. The call uses new guidance figures, a new initiative name announced in the quarter, and three new analyst names who haven't appeared on previous calls. All of these produce caption errors that would have been avoided by a pre-call glossary update triggered by the IR preparation process. The fix is a quarterly glossary update workflow owned by the IR team, triggered when earnings call talking points are finalised.

6. M&A announcement video published before caption team is notified

The M&A announcement is public at 6:30am. The CEO video message goes live on the intranet at the same moment. The caption team learns about it from the same internal communications email that all employees receive. By the time a caption request is submitted, reviewed, processed, and returned, the video has been live for several hours without captions. The fix is a pre-announcement protocol: the Corporate Communications team notifies the caption programme manager under NDA at the same time the communications materials are finalised, so that the caption can be produced and reviewed in advance and attached to the video at the moment of the announcement.

7. Investor day recordings archived without caption review

Investor day is an exhausting, all-day event. By the time the recordings are edited and uploaded to the IR website, the team responsible for the event has been running for 12–14 hours. Caption upload is the last thing anyone is thinking about. The recordings go up without caption tracks. Two weeks later, someone notices. By that time the recordings have been viewed hundreds of times by analysts and investors with no caption access. The fix is to include investor day caption production in the investor day planning budget and timeline, not as an afterthought after the event — and to designate a specific person as caption upload owner who is not also responsible for 12 other investor day deliverables.

8. The Board materials captioning gap

Board meeting recordings and board committee recordings are hosted in a board portal that does not support caption file upload, or that does support it but no one has set up the caption track because the board portal is managed by the Corporate Secretary's office and the caption programme is managed by L&D or Corporate Communications, and neither team knows the other's platform well enough to set up the integration. Board materials captioning is the responsibility of no single team, and the board portal's caption support (or lack thereof) is unknown to the caption programme. The fix is a one-time audit of the board portal's caption capabilities and a handoff protocol between the Corporate Secretary and the caption programme for any recorded board content.

FAQ

Does a private company need to caption its all-hands recordings and investor communications?

Private companies with 15 or more employees are covered by ADA Title I and must provide effective communication to employees with disabilities in all employment contexts, including all-hands recordings and CEO video messages distributed to employees. ADA Title I applies to private employers regardless of public company status. Private companies that raise venture capital or private equity typically have investor calls or board update videos, but these are not public webcasts — they are distributed to a specific group of investors, not to the general public. Private company investor communications are not ADA Title III content in the same way as public company earnings calls. However, if a private company's investor communications are accessible to the general public (e.g., a public webinar open to any attendee), they would be Title III-covered. The Title I obligation for employee communications content applies regardless of company stage or investor type.

Our earnings call transcript is produced by a professional transcription service. Does that count as a caption track?

No. A professional earnings transcript is a flat text document without timestamps. WCAG SC 1.2.2 requires synchronised captions — captions that are time-aligned to the audio and displayed as the corresponding audio plays. A flat transcript is not synchronised to the audio; it is a separate document that a viewer must read alongside the video. The professional transcript is an excellent source material for producing a caption track because its accuracy is high, but it must be converted to a timed SRT or VTT file with timestamps aligned to the recording before it qualifies as a synchronised caption track. The conversion from professional transcript to timed caption file typically takes 30–90 minutes for a 60-minute earnings call and can be built into the post-call IR workflow. See the 48-hour accuracy window section for the workflow design.

What should we do about the earnings call recordings that are already on our IR website without caption tracks?

The back-catalogue of uncaptioned earnings recordings represents a compliance gap that can be addressed systematically. For WCAG 2.1 AA compliance, all pre-recorded audio-visual content on your website must have synchronised captions — including historical recordings. The prioritisation approach for back-catalogue remediation: first, the current quarter's recording (if it does not have a compliant caption track, this takes priority over everything else); second, the past four quarters' recordings (the most likely to be accessed by analysts and investors tracking recent performance); third, the remaining back-catalogue in reverse chronological order. Each historical earnings recording will already have a professional transcript if the company used a transcript service — that transcript is the source material. The conversion cost per recording (transcript to timed caption file) is lower than producing a caption from scratch and can often be processed in batches. Budget planning guidance covers back-catalogue cost modelling.

Our CEO uses a teleprompter for all-hands. Can we use the teleprompter script as the caption source?

Yes, with caveats. A teleprompter script is a word-for-word text that the CEO reads verbatim — the most accurate text source for the prepared remarks portion of the recording. The caption production workflow for a teleprompter-based recording: obtain the final teleprompter script from the communications team, align it to the recording audio using a caption timing tool (this adds timestamps to the text), review for any off-script deviations (the CEO may have ad-libbed, corrected a mis-speak, or improvised), and produce a timed SRT or VTT. The caveat is the Q&A portion: the teleprompter script does not cover Q&A, which is unscripted. The Q&A section requires a separate caption source — either ASR-generated with glossary and review, or transcribed from the audio. The combined approach (teleprompter script for prepared remarks + reviewed ASR for Q&A) is the most efficient workflow for CEO all-hands recordings with a scripted prepared section followed by an unscripted Q&A.

What is the most common investor relations platform incompatibility with caption files?

The most common compatibility problem is SRT versus VTT format mismatch. Most caption files produced by ASR systems or professional captioning services can be exported in either SRT or VTT format — the two formats contain the same content but use different syntax. Q4 Inc. and Notified both accept VTT; Wistia accepts both SRT and VTT; Vimeo accepts both; YouTube accepts SRT, VTT, and several other formats. If the IR platform or hosting platform rejects a caption file, check the format: upload the VTT version if SRT was rejected, or vice versa. The second most common compatibility problem is character encoding — caption files must be UTF-8 encoded, not ASCII or ISO-8859-1. If special characters (the ampersand (&), non-breaking spaces, or non-ASCII currency symbols) appear in the caption text, the encoding must be UTF-8 for the file to upload correctly. Most professional captioning services produce UTF-8 files by default; caption files converted from Word documents or emailed text may not be.

How do we handle the CEO's earnings commentary that is also included in an internal employee-communication recap?

Content that appears in both an external investor context (IR website) and an internal employee context (intranet) must meet the higher of the two compliance standards for each audience. In practice, WCAG SC 1.2.2 applies to both ADA Title I (employees) and ADA Title III (public), so the accuracy standard is the same — 99% for pre-recorded content. The caption track produced for the IR website version is the correct track for the internal version as well; it does not need to be produced twice. The workflow consideration is coordination between the IR team that manages the external version and the Corporate Communications team that manages the internal recap — both need to use the same caption track. If the internal version is edited differently from the external version (e.g., the earnings guidance portion is removed for the internal recap, or B-roll is added), the caption track must be re-aligned to the edited version rather than copied from the external version without verification.

Where does executive communications captioning fit in the broader accessibility programme priority ranking?

Executive communications captioning — and specifically earnings call and investor day captioning — should be treated as a Tier 1 programme priority alongside the training content that L&D already captions, rather than as an extension of lower-priority communications captioning work. The rationale for Tier 1 treatment: external investor-facing content carries public accommodation (Title III) obligations with no exemption for company size; the content is actively used by a sophisticated investor and analyst audience who notice quality problems; the financial vocabulary failure modes are severe enough to produce materially misleading captions; and the cost of a disability accommodation request and response cycle is substantially higher than the cost of a prevention programme. The ROI framing for finance and executive audiences for accessibility investment applies directly to this content category — the risk management argument (preventing an investor-facing accessibility failure that produces reputational and legal exposure) is the most compelling argument for investor relations and C-suite sponsorship of an executive communications caption programme.