Something unusual is happening in the real-time conversational AI space. Four separate companies — founded independently, in different cities, by different teams, with different investors — have all arrived at the same word to describe what they do: cuecard. And a fifth signal has emerged from an entirely different direction: a major enterprise software company independently chose "cue cards" as the literal name for its flagship real-time coaching feature, years before any of the startups existed.

There is cuecard.ai, an Austin-based sales enablement platform founded in 2022. There is cuecard.dev, a developer-focused real-time assistance tool. There is cuecardai.com, which bills itself as a real-time battle card platform offering AI summaries and objection handling for sales calls. And there is usecuecard.com — registered in March 2026 and currently in active pre-launch development, operating a waitlist while inviting early users in small batches. Then there is the fifth signal: Clari Copilot — formerly Wingman, a platform serving over 5,000 enterprise customers — which has used "cue cards" as the literal product name for its real-time sales coaching feature for years. Clari did not name itself after the concept. It named its most important feature after it. That is a different kind of validation entirely.

None of these companies coordinated. None of them copied the others. The startups looked at what they were building and reached for the most intuitive word available. Clari's product team, building enterprise software years earlier, reached for the same word independently. That is the convergence signal.

What makes "cuecard" so apt

The appeal of "cuecard" as a category name is not arbitrary. It earns its status through three qualities that are rare to find in a single compound word. First, it is universally understood. Anyone who has watched a television broadcast, given a speech, or sat in a school classroom knows what a cue card does — it gives you the right words at the right moment. No explanation required. Second, it is descriptively precise. A cue card doesn't think for you; it prompts you. That distinction matters in a category where the technology is explicitly about augmenting human conversation rather than replacing it. Third, it carries no negative connotations in its professional sense. Unlike words such as "cheat" or "overlay" or "stealth" — all of which competitors have used — "cuecard" is universally positive in professional contexts. Actors use them. Presidents use them. News anchors use them. The word is associated with preparation and professionalism, not deception.

The workaround names are the tell

The most revealing evidence is not the companies that chose "cuecard" directly — it is the form their alternatives took. usecuecard.com did not choose a different word entirely. They chose "usecuecard" — the "use" prefix being the classic startup workaround when the exact .com is unavailable. Similarly, cuecardai.com did not find another name; they appended "ai" to the word they actually wanted. These are not alternative naming strategies. They are constraint strategies — what founders do when the name they want is taken and they are not yet positioned to acquire it.

This pattern is well established in technology naming history. Uber launched as UberCab. Loom launched on useloom.com before acquiring loom.com. Zoom operated on zoom.us for years before acquiring zoom.com. In each case, the workaround name was a holding position — a temporary settlement while the company built enough value to justify acquiring what it actually wanted.

What convergence predicts

Category naming convergence is a leading indicator, not a lagging one. It happens before a category is fully formed, when the concept is clear enough to name but not yet dominant enough for one company to have locked in the terminology. That moment — right now, in the first half of 2026 — is precisely when category names get decided.

The companies and investors paying closest attention to this space should be watching the language as carefully as they watch the technology. The word that sticks will shape search behavior, media coverage, investor framing, and ultimately market structure for the next decade. Four independent startups and one major enterprise product team have already voted — without knowing about each other. The question now is who owns the word they all voted for.

In April 2025, two Columbia students turned a side project into a company, raised $5.3 million in seed funding, and within six weeks closed a $15 million Series A led by Andreessen Horowitz. By the end of 2025, Cluely had reportedly reached $20 million in annual recurring revenue. The product: a real-time AI overlay that listens to conversations and surfaces prompts during live calls, meetings, and interviews. The result: one of the fastest early trajectories in recent AI startup history.

Cluely's achievement is real. They identified a genuine human need — the anxiety of high-stakes conversations — and built something that addresses it directly and visibly. Andreessen Horowitz doesn't lead Series A rounds on bad bets. The category they helped bring into mainstream awareness is important, growing fast, and worth watching carefully. What is equally worth examining is the specific market segment that Cluely's positioning leaves open.

The enterprise buying dynamic

Enterprise software procurement operates on a different logic than consumer adoption. Where consumers can download and use a product in minutes, enterprises require security reviews, compliance assessments, and vendor evaluations that can take months. The questions procurement teams ask are predictable: Does the vendor hold SOC 2 Type II certification? How is user data stored and who has access to it? What happens in the event of a breach? Can the tool be deployed without violating wiretapping laws in the relevant jurisdictions? A 2026 survey found that 78% of enterprise SaaS buyers treat SOC 2 Type II as a non-negotiable requirement before procurement conversations begin.

These questions are not unique to Cluely — they apply to every tool in this category. What matters is how a company's positioning affects its ability to answer them convincingly. An enterprise security team evaluating a tool that brands itself around stealth and invisibility faces an inherent tension: the product's stated value proposition is that nobody knows it's there. For a consumer user in a job interview, that's a feature. For an enterprise deploying the tool across a sales team, it raises questions about consent, disclosure, and legal liability that are genuinely difficult to resolve. Several large technology companies have responded to this by explicitly banning screen-scraping overlay tools from corporate devices altogether.

The gap this creates

The gap is not about Cluely specifically — it is structural. The consumer and enterprise markets for real-time conversational AI assistance want related but meaningfully different things. Consumer users want a discreet personal advantage in competitive moments. Enterprise buyers want a transparent, auditable, compliance-ready tool that improves team performance at scale. The technology underlying both use cases is similar. The product positioning, security architecture, and brand language required to serve them are quite different.

Enterprise players are already moving to fill this space from their existing positions. Clari Copilot has been selling real-time "cue card" functionality to sales teams for years. Gong, HubSpot's Breeze Copilot, and Salesforce's Agentforce all shipped real-time coaching features in 2025. The AI coaching platform market is projected to reach $11.1 billion by 2035. The enterprise segment of that market — large, well-funded, and actively looking for compliant solutions — is served today by partial solutions from established players rather than a purpose-built tool with a clean brand identity.

The naming dimension

There is a branding observation worth making alongside the market one. The enterprise-appropriate vocabulary for this category clusters naturally around words like "copilot," "assistant," and — with notable frequency across the startups in this space — "cue card." The word carries professional connotations that enterprise buyers find comfortable: preparation, structure, expertise made accessible. It is the same word that four independent startups have reached for independently, and the same concept that Clari Copilot named its flagship enterprise feature years before those startups existed. That convergence is not accidental. It reflects something real about how the market wants to think about this technology.

Every major technology category goes through a naming period — a window, usually measured in years rather than decades, during which the terminology is still fluid and the dominant descriptor has not yet been decided. We are in that window right now for real-time conversational AI assistance. The category is real. The technology works. The market is growing fast. What we don't yet have is a name that everyone agrees on.

This matters more than it might seem. Category names are not just marketing labels — they are infrastructure. They determine what people search for, how analysts frame the market, what journalists write about, how investors categorize their portfolios, and ultimately which company gets positioned as the defining player. The company that names a category often ends up owning it.

The current naming landscape is fragmented

Scan the space today and you find at least half a dozen descriptors competing for dominance. "Conversation intelligence" was the early enterprise term, used by Gong, Chorus, and Clari — but it describes post-call analytics more than real-time assistance. "Real-time coaching" is more accurate but long and functional, better suited to a feature description than a category name. "AI copilot" has been colonized by Microsoft's Copilot branding and now risks becoming a generic modifier rather than a category name. "Meeting assistant" describes the platform rather than the capability. "Stealth assistant" is a framing that enterprise procurement teams tend to evaluate carefully, given the consent and compliance questions it raises.

None of these names has won. Each describes part of the category without capturing the whole. And into that gap, a more intuitive word keeps appearing.

How category names actually get decided

It is tempting to think that category names are decided by the dominant company — that whoever wins the market gets to name it. History suggests the opposite. "Search" as a category name predates Google's dominance. "Social media" was in common use before Facebook became the defining platform. "Cloud computing" emerged from user and analyst language before any single company claimed the category.

Category names are decided by usage — by the cumulative weight of founders naming their companies, users describing their experience, and journalists reaching for a shorthand. They emerge from the bottom up, and they solidify surprisingly quickly once they start. The window during which a category name can be influenced is narrow. Before the name is settled, anyone can shape it. After it is settled, it belongs to everyone and no one.

The case for "cuecard" as the category name

The strongest argument for "cuecard" as the category descriptor is not its elegance — though it is elegant. It is its behavioral evidence. Four independent startups have already reached for it. Clari Copilot named its real-time sales coaching feature "cue cards" years before those startups existed. Patent filings at Microsoft reference "feature cue cards" as a named technical concept. The word has been independently validated by founders, product teams, and engineers across multiple organizations and time periods. That is not a word being pushed by a marketing team. That is a word being pulled toward a category by the gravity of its own aptness.

The question is not whether "cuecard" is the right name for this category. The evidence suggests it already is, at least in the informal vocabulary of the people building and using these tools. The question is whether the formal vocabulary — the analyst reports, the investor theses, the media coverage, the product names of the major players — catches up before or after the window closes.

The cue card is one of the oldest communication tools in professional use. Long before artificial intelligence, before computers, before even the teleprompter, performers and speakers used physical index cards to navigate complex conversations — a few words on a card held just out of sight, prompting the next thought at precisely the right moment. What is new in 2026 is not the concept. It is the intelligence behind it.

Understanding where the cue card came from helps clarify where it is going — and why this particular moment in technology history is generating so much activity around the idea of real-time conversational prompting.

The physical era: index cards and autocue

The earliest cue cards were literal cards — index cards or sheets of paper held by a crew member just below or beside the camera, out of audience view, carrying the key points a presenter needed to make. Television production industrialized this into the teleprompter, a device that scrolls prepared text in front of a camera lens so a presenter can appear to speak extemporaneously while reading verbatim. The teleprompter solved the memory problem for prepared remarks. What it could not solve was the unscripted conversation — the sales call, the negotiation, the interview, the meeting where no script exists and the right response depends on what the other person just said.

The digital transition: battle cards and knowledge bases

The first digital evolution of the cue card appeared in sales enablement software in the late 2010s. Wingman, founded in 2018 and later acquired by Clari, was one of the earliest companies to explicitly use the term "cue cards" as a product feature name — describing the automatically surfaced talking points that appeared on a sales representative's screen when a competitor was mentioned or an objection was raised. These digital cue cards were manually authored, keyword-triggered, and entirely text-based. They were smarter than index cards but still essentially static: a human had to write the content in advance, and the system simply retrieved it at the right moment.

This era produced what the industry called "battle cards" — a militaristic framing that reveals the competitive anxiety behind the feature. The knowledge was there, delivered at the right moment, but the intelligence was human, not artificial.

The AI era: real-time generation

The transition from retrieval to generation changed everything. When large language models became capable of understanding conversational context in real time and generating relevant responses in under 400 milliseconds — a threshold the category crossed in 2025 — the cue card stopped being a lookup system and became a thinking system. The difference is profound. A retrieved cue card can tell you what you prepared to say about a competitor. A generated cue card can tell you what you should say about this specific objection, from this specific prospect, in the context of this specific conversation, right now.

Microsoft's patent US11902690B2, granted in February 2024, describes precisely this transition. The system displays a teleprompter transcript on screen, uses a first machine learning model to convert the presenter's live speech to text in real time, and a second model to determine where in the transcript the presenter currently is — then automatically scrolls the cue text in sync with their voice. That is a digital cue card driven by live speech recognition, filed under Microsoft Technology Licensing LLC. Microsoft also holds patents covering proactive intelligent assistants that surface "cue" recommendations during live conversations without requiring an explicit request — the same shift from passive retrieval to active generation, applied to messaging and collaborative applications rather than presentations. The technology company with arguably the deepest investment in this space has been thinking in cue card terms for years.

What comes next

The trajectory from index card to teleprompter to battle card to AI overlay follows a consistent logic: each generation reduces the gap between what a speaker knows and what they need to say, while increasing the intelligence of the assistance provided. The next step is not difficult to project. Real-time AI cue cards will become ambient — present in every professional conversation, transparent to all participants, integrated into the communications infrastructure that organizations use daily. The question of whether this technology becomes ubiquitous is largely settled. The question of what we call it when it does is still open.

On March 7, 2026, a company registered usecuecard.com and began building a real-time conversational AI platform on it. The product is described as "a private AI copilot for interviews, meetings, presentations, and sales calls." The platform is currently in pre-launch, operating a waitlist while inviting early users in controlled batches. The team is clearly building something with ambition and structure. What is worth examining is not the product — it is the name. Specifically, it is the prefix.

"Use" is a workaround. In startup domain naming, the prefixes "use," "get," "try," and "my" have a specific cultural meaning: they signal that the founder wanted the unprefixed domain and found it unavailable. They are the naming equivalent of a consolation prize — functional, recognizable, but understood by everyone in the industry as a second choice. The company that registers usecuecard.com is telling you, in the language of domain naming, that they wanted cuecard.com. The prefix is not a creative choice. It is a constraint.

The history of workaround names

This pattern is well documented in technology history. Uber launched as UberCab before acquiring the cleaner brand. Loom launched on useloom.com before acquiring loom.com — a transition that required significant capital and negotiation once the company had grown valuable enough to justify it. Zoom operated on zoom.us for years, the country-code domain serving as a workaround while the company built the revenue and relationships needed to acquire zoom.com. In each case, the workaround name served its purpose — but it was always understood internally as temporary.

What workaround names reveal about category formation

The more interesting insight is not about any individual company — it is about what the cluster of workaround names in a given space reveals about category formation. When cuecardai.com, usecuecard.com, cuecard.ai, and cuecard.dev all exist simultaneously, all building products in the same category, all having chosen variants of the same word, it means the category has a name. It just doesn't have an owner yet. The market has reached consensus on the right word before any single company has secured the authoritative form of it. That is a specific and unusual condition — and historically, it precedes a period of rapid consolidation around whoever does secure that authoritative form.

The upgrade imperative

For any funded startup operating on a workaround domain, the math changes as the company grows. In the early stages, a workaround name is a reasonable trade-off — the capital required to acquire the clean domain is better spent on product and people. But as a company raises successive rounds, builds brand equity, and begins thinking about enterprise sales cycles, the workaround name becomes an increasingly visible liability. Enterprise buyers notice. Press coverage reads awkwardly. The domain shows up in every email signature, every slide deck, every sales call. The cost of the workaround, measured in brand friction, compounds with scale.

The upgrade from a workaround name to the clean domain is not a question of if — for any company that succeeds in this category, it is a question of when. And the longer a company waits, the more brand equity it builds on the workaround name, making the eventual migration more disruptive and more expensive. The optimal time to upgrade is before the name becomes load-bearing. That window, for the companies currently operating in the real-time conversational AI space, is open right now — and it will not stay open indefinitely.

In July 2025, the Pickle team built something in four days that went viral across the developer community. They called it Glass — an open-source desktop assistant that lives invisibly on your Mac, watches your screen, listens to your audio, and surfaces real-time suggestions during meetings without ever appearing in a screenshot, a screen recording, or your Dock. Within days it had accumulated nearly 2,000 GitHub stars. The description its creator used was precise: "the open-source Cluely."

Glass is software. But the Pickle team's ambition didn't stop there. In January 2026, they announced Pickle 1 — a pair of AI-powered smart glasses, priced at $799, targeting Q2 2026 delivery, with a full-color AR display and what they called a "soul computer" operating system designed to remember everything you see and hear. The jump from a desktop overlay to a wearable is significant. It represents the logical endpoint of a trajectory that started with the physical index card and has been moving, generation by generation, closer to the moment of the conversation itself.

From overlay to wearable

The screen overlay model — Cluely, Glass, usecuecard.com, LockedIn AI — solves one problem while creating another. It puts the cue card in your field of vision, but only if you're looking at a screen. For a job interview conducted over video, or a sales call in a browser, the screen is already occupied. The overlay competes for visual attention. It is useful but awkward — a workaround, not a native experience.

Wearable AI addresses the awkwardness directly. Smart glasses can display information in your peripheral vision without occupying the screen at all. The cue card appears where it has always conceptually belonged — just out of sight, just in time, invisible to everyone except you. Even Realities launched a feature called Prep Notes in March 2026 for exactly this purpose: real-time conversational cues delivered through smart glasses, with their CEO describing it as "something that shows up with you in the moments that matter — starting with the conversations happening every day." That is a cue card. It is just worn, not displayed.

The open-source signal

Glass's rapid adoption as an open-source project is worth examining separately from the hardware question. Glass reached 1,800 GitHub stars and 323 forks quickly, with a highly active developer community. Open-source adoption at that pace signals something specific: developers want this technology, they want it free, and they want to build on it. The open-source model also bypasses the enterprise procurement concerns that make Cluely difficult to deploy at scale — a self-hosted, auditable, locally-run version of the same capability is a fundamentally different compliance conversation.

The combination of open-source software and emerging wearable hardware suggests the category is about to fragment in an interesting way. Consumer and developer users will likely move toward wearables and open-source tools. Enterprise users will move toward transparent, compliant, cloud-managed platforms. Both are cue cards. Both serve the same fundamental need. The question of what to call the category becomes more important, not less, as it diversifies across form factors.

What this means for the category name

There is something almost poetic about the trajectory. The physical cue card — a literal card, held by a human, just out of frame — is migrating through software into hardware and arriving, eventually, as an augmented reality overlay in a pair of glasses. The form factor changes completely at each step. The concept does not change at all. It is always the same thing: the right words, in the right moment, visible only to you.

A category that spans open-source desktop software, enterprise SaaS platforms, consumer subscription apps, and wearable AR hardware needs a name that is form-factor agnostic. It needs a word that means something whether the technology runs on a MacBook or sits on your nose. "Cuecard" is that word. It describes the function, not the form. And as the form continues to evolve — as it will — the function remains constant.

While consumer-facing tools like Cluely have dominated the conversation about real-time AI assistance, a parallel and substantially larger market has been building quietly in the enterprise for nearly a decade. Contact centers — the industrial-scale conversation environments where agents handle millions of customer interactions daily — have been deploying real-time AI coaching systems since 2017. The technology is mature, the results are documented, and the market is enormous. What this segment of the category lacks is the same thing the consumer segment lacks: a common name.

Cresta is the most instructive example. Founded in 2017 and born out of the Stanford AI Lab, Cresta has raised $282 million across eight funding rounds and reached a $1.6 billion valuation, with $52 million in annual recurring revenue and 668 employees as of 2026. Its investors include Andreessen Horowitz, Sequoia Capital, Greylock Partners, and — notably — Accenture and Workday Ventures, signaling deep enterprise credibility. Its customers include United Airlines, Cox Communications, and Porsche. Cresta operates as an AI copilot that sits on every customer interaction in a contact center, listening to phone calls, chats, and emails in real time, turning real-time insights into real-time actions. That is, functionally, a cue card system operating at enterprise scale.

The contact center as category laboratory

The contact center matters for understanding this category not because it is the most glamorous application of real-time AI assistance — it is not — but because it is the most data-rich. Cresta, NICE CXone, Genesys, and Balto have collectively deployed real-time coaching systems across millions of agent interactions and accumulated years of outcome data. They know which prompts work. They know when to surface a suggestion and when silence is more effective. They have solved, in a controlled enterprise environment, the hardest technical and behavioral problems that consumer tools are only beginning to encounter.

The performance gap the technology addresses is well documented. The gap between top and bottom agents in the same contact center can reach 230%, according to McKinsey's analysis of contact center sales performance. Traditional coaching — feedback delivered days or weeks after a conversation — cannot close that gap because by the time an agent sits down for a session, the context has faded. Real-time coaching closes it by turning every live conversation into a coaching moment. Less than half of contact center agents report receiving effective on-the-job coaching today, and the gap shows up directly in business results. The enterprise demand for this technology is not speculative. It is documented and growing.

The vocabulary problem

Despite years of development and hundreds of millions in investment, the enterprise segment of this category has not settled on a name. The platforms in this space use a patchwork of descriptors: "real-time agent coaching," "behavioral guidance," "live hints," "agent assist," "next-best-action prompts," "AI copilot." Each describes the same fundamental capability from a different angle. None has become the category name.

The consumer side of the category has moved faster toward a common vocabulary — partly because it is newer and therefore less entrenched in legacy terminology, and partly because the word "cuecard" has emerged naturally from multiple independent starting points. The enterprise side is converging on the same concept through a different path: not from founders naming startups, but from product teams describing features. Cresta's "behavioral guidance." NICE's "real-time interaction guidance." Genesys's "agent assist." All are cue cards. None are called that.

Why the enterprise and consumer markets will converge

The current separation between enterprise contact center coaching tools and consumer real-time AI assistants is a product of history rather than technology. Enterprise tools were built for contact centers because that is where the regulated, auditable, high-volume use case justified the investment. Consumer tools were built for individuals because generative AI made personalized real-time assistance economically viable at the consumer level for the first time.

The underlying technology is identical. An AI system that listens to a live conversation, understands context, and surfaces the right prompt at the right moment works the same way whether the user is a contact center agent handling a billing dispute or a sales representative on a first call with a prospect. The enterprise and consumer markets are not different categories — they are different deployment contexts for the same category. And as enterprise tools expand beyond contact centers into general sales, HR, and executive communications, and as consumer tools build enterprise-grade compliance and security, the two contexts will converge around a common vocabulary.

The category name that bridges both contexts needs to work for a United Airlines contact center agent and for a job candidate in a video interview. It needs to be professional enough for enterprise procurement and intuitive enough for a consumer download page. It needs to describe the function without prescribing the form — whether the prompt appears on a screen, in a headset, or in a pair of smart glasses. "Cuecard" satisfies all of those requirements. The enterprise has been building toward it for nearly a decade without knowing that was the name.

There is a reliable signal that a technology category has moved from speculative to strategic: the largest companies in the world stop building and start buying. In the second half of 2025, two of them did exactly that. Amazon acquired Bee, an always-on wearable that listens to a user's day and turns it into searchable summaries, in July 2025. Five months later, in December 2025, Meta acquired Limitless — a conversation-recording pendant backed by Sam Altman and Andreessen Horowitz that had raised more than $33 million. Within a single year, two of the most valuable companies on earth made acquisitions in the same narrow space: AI that listens to live conversations and assists in real time.

When Meta and Amazon move in the same direction within months of each other, it is no longer a trend. It is a category.

What the acquirers actually bought

It is worth being precise about what changed hands. Limitless built a pendant that recorded the conversations around its wearer and turned that audio into transcripts, meeting recaps, and a searchable record of the day. Bee built a screenless, always-on wearable that listens and constructs a running context of what the user has been doing, who they have been talking to, and what is coming up. Both companies solved the same fundamental problem from the same direction: capture the conversation, understand it, and hand back the right information at the right moment. That is the cue card concept, rendered in hardware.

Meta's ambitions for what it acquired are not modest. Internal plans point to a business subscription called "Wearables for Work" and a target of roughly 10 million wearable units in the second half of 2026. This is not an experiment. It is a platform bet at consumer scale, aimed squarely at the professional market — the same market that the enterprise coaching platforms and the consumer copilots are both pursuing from their own directions.

The convergence becomes physical

For seven previous installments, this publication has tracked a category converging on a concept from multiple directions — consumer copilots, enterprise coaching platforms, meeting assistants, and the naming convergence around the word "cuecard." The wearable acquisition wave adds the final dimension: hardware at the scale of the largest technology platforms in the world. The cue card is no longer just a screen overlay or an enterprise feature. It is becoming something you wear, backed by the balance sheets of Meta and Amazon.

This matters for the category's vocabulary because hardware forces clarity. A software feature can hide behind a product name. A wearable device that delivers real-time conversational assistance to millions of professionals needs a category — a word that describes what the entire class of products does, independent of which company makes it or what form it takes. Search engines need it. Retailers need it. Journalists need it. The category that Meta and Amazon are now investing in at scale does not yet have that word in common use. But the candidates are few, and one of them keeps appearing.

The consent problem reinforces the naming problem

The wearable wave also sharpens a tension this publication has noted before. A device that records everyone within earshot raises consent questions that a screen overlay does not. A microphone pointed at every conversation pulls in bystanders who never agreed to be recorded, and the legal questions around this are already producing litigation. The companies that win at scale in this category will be the ones that solve for trust — transparent capture, clear consent, professional positioning.

That requirement favors a particular kind of brand. A category built on always-listening hardware cannot afford a name that connotes surveillance, stealth, or deception. It needs a name that signals preparation and professionalism — the opposite of secrecy. "Cuecard" carries exactly that connotation. It describes a tool that helps you say the right thing, openly, the way a teleprompter helps a broadcaster or an index card helps a presenter. As the category moves into always-on wearables and the trust question becomes central, the professional, non-stealth associations of the word "cuecard" become not just appropriate but strategically necessary.

When the largest companies in the world start acquiring in a category, they eventually need to name what they have built. The naming window that this publication has tracked since its first installment is the same window Meta and Amazon are now operating inside — whether they realize it yet or not.

On December 15, 2025, Gartner published its first-ever Magic Quadrant for Revenue Action Orchestration — a new category, evaluated across twelve vendors, defined in Gartner's own words as emerging "from the convergence of previously separate sales engagement and revenue intelligence markets." Gong, Clari, and Outreach were named Leaders. Salesloft was named a Visionary. For an industry analyst firm to publish a first Magic Quadrant is a specific kind of event: it means enough capital, enough customers, and enough vendor activity has accumulated in one place that the market can no longer be described using the old category boundaries. Gartner does not name categories speculatively. It names them after the evidence is already in.

The report did something else worth noting separately from the category name itself. Its accompanying Critical Capabilities analysis evaluated vendors across four specific use cases, one of which was "coach sales talent." Gong ranked first in all four. That is a formal, third-party statement that real-time coaching is not a feature bolted onto a category — it is one of the defining capabilities of the category itself.

What "Revenue Action Orchestration" describes, and what it doesn't

"Revenue Action Orchestration" is a real category name, and it is already doing useful work — it gives analysts, investors, and buyers a shared term for the market as a whole. What it does not do is describe the specific moment this series has tracked since its first installment: the point inside a live conversation where a system decides what a person should say next, and shows it to them before the conversation moves on. RAO is the name for the market. It is not the name for the mechanism. Those are different naming problems, and only one of them has been solved.

This is a familiar pattern. "Cloud computing" named a market years before "container" or "serverless" named the mechanisms inside it. "Search" named a market before "crawler" and "index" became the working vocabulary of the people building it. Category-level names and mechanism-level names solve different problems and tend to arrive on different schedules. Gartner has now solved the first problem for this market. The second remains open, and the evidence gathered across this series continues to point toward the same word.

What else has moved since the last installment

The Seismic–Highspot merger, described in the eighth installment as an announced deal, closed on August 18. The combined company operates under the Seismic name, serves 2,500 customers and 3.5 million users, and its CEO has publicly framed the moment as "a fundamental shift from GTM preparation to GTM performance" — language that describes, again, the shift from preparing for a conversation to acting inside one. Zoom acquired Common Room in July, its second acquisition in nine months in adjacent territory, and shipped its own real-time coaching feature to general availability the same month. Two more independent products, unconnected to each other or to any company previously named in this series, have separately reached for "cue card" language in their own marketing since the eighth installment published. The naming convergence this series set out to document has not slowed down since the retrospective. If anything, the rate of independent confirmation has increased.

Why the gap matters more now, not less

A newly formalized category with an unsettled mechanism-level vocabulary is a specific and narrow window. Before Gartner's report, a skeptic could reasonably argue the entire category was still speculative — too early to be sure a real market existed at all. That argument is no longer available. What remains open is only the naming question this series has tracked from the beginning, and formal category recognition tends to accelerate naming convergence rather than settle it, because it is precisely the moment when analysts, journalists, and investors start needing a shorthand for the thing Gartner just validated. The window in which that shorthand gets decided is usually short. It is open right now.

Over the summer and into September, three of the platforms already covered in this series each published product or roadmap updates that leaned on the same descriptive word: real-time. None of them reached for a name.

In July, Zoom made Sales Assist, which it describes as real-time deal guidance, generally available. On September 15 it went further, introducing what it called an AI-powered "revenue OS" — a unified system connecting buyer intelligence, customer conversations, and revenue execution, anchored by two new capabilities, Engage and Forecast, and the general availability of Common Room by Zoom. The same week, IDC named Zoom a Leader in its 2026 MarketScape for unified revenue orchestration — a second analyst firm, following Gartner's Magic Quadrant covered in the previous installment, formally recognizing a market spanning conversation, forecasting, and execution.

A Gartner Visionary names its own next stage "real-time"

Nine days later, on September 24, Salesloft went further than a routine product update. The company — a Gartner Visionary, and now operating as a single entity with Clari following the two organizations' September 2 brand consolidation — described the next stage of its roadmap under a heading it chose for itself: "Expansion of Real-Time CI Capabilities." Its own language: Salesloft's Conversation Intelligence "already listens and acts in real time on every call, delivering live coaching and transcriptions," and the next step is "enabling agentic workflows to respond during the conversation." CEO Steve Cox went a step further, listing "Real-Time Conversation Intelligence" as one of the four named pillars of the roadmap. That is, in the company's own words, a description of the exact mechanism this series has tracked since its first installment — with no name attached to it beyond a real-time version of a term the industry already uses for something else.

Salesforce approaches from the capture side

Salesforce came at the same territory from the capture side. Momentum — the conversation-intelligence product Salesforce announced it was acquiring in February — was featured at Dreamforce (September 15–17). Salesforce describes it as capturing every call, email, and meeting and writing that data back into Salesforce "in real time," with production availability expected later this quarter. Here "real time" means how quickly data reaches the CRM, not guidance delivered while the customer is still talking — which is itself the point. The phrase is being stretched across several different meanings, and none of them is a name for the live layer.

The smaller end of the market is funded too

This is not only an incumbent story. In February, Winn.ai — a roughly 40-person company with customers including Deel, Intercom, and Snyk — raised an $18M Series A to bring real-time AI guidance to sales teams, more than double the funding of Othello, the entrant covered in the previous installment. That round came about two months after Gartner named the market, and it shows fresh venture capital aimed squarely at the mechanism this series has tracked, not just at the market Gartner defined around it.

What's still missing

None of this changes the conclusion of the ninth installment. It reinforces it. Gartner named the market in December. Nobody has yet named the mechanism inside it. Even Gong, ranked first in the "coach sales talent" capability, names its products by function — Engage, Forecast, Enable, Orchestrate — rather than by any category term for the capability itself. Three platforms keep saying "real-time," in three different senses. Not one of them has said what to call it.