▶️ Listen to this episode
[ Apple Podcasts ] [ YouTube ]
David Beckham speaking nine languages was never the business.
It looked like one, for a moment. The 2019 campaign with Malaria No More UK put Synthesia on every marketing trade publication’s radar. Synthetic media had arrived, and it had a famous face attached.
But I’ve seen this pattern before in my own work: the flashiest campaign in a company’s history is sometimes the clearest sign its business model is broken.
That’s exactly what happened here.
The Service Trap Disguised as a Product
Selling AI dubbing to media and advertising clients felt like software. It wasn’t.
Every project required legal sign-off, custom creative direction, and long production timelines. The AI did a sliver of the work. The rest was the same friction that has always slowed down media production.
For agencies and broadcasters, synthetic video was nice to have. Not essential.
I’ve watched plenty of founders fall in love with their own demo. The demo gets headlines. It rarely gets recurring revenue.
Finding the Real Pain Point
Synthesia’s founders eventually looked past the spectacle and found something far less glamorous: corporate training videos.
Companies had a massive backlog of content that needed to be video—onboarding, compliance, sales enablement—but stayed trapped in PDFs and slide decks. A professional video could cost tens of thousands of dollars per finished minute. Most teams simply couldn’t justify it.
That’s a painkiller, not a vitamin. And painkillers are what you build a company around.
This is the part of ai video marketing strategy that gets overlooked. It’s not about finding where your technology looks impressive. It’s about finding where the alternative to your product is nothing at all.
Betting Against Investor Consensus
The technical spark came from a 2017 research paper on neural-network-generated video, the kind of academic work most operators never read, let alone act on.
Roughly a hundred investors passed. Synthetic humans sounded fringe.
Then a cold email landed in front of Mark Cuban, who’d already experimented with the same underlying research. No phone call, just a long email thread—and a $1 million check.
I bring this up not as a feel-good footnote, but because it says something about category-creation: the people closest to a genuinely new technology are often more convinced by it than the people whose job is supposedly to spot the future.
Choosing Reliability Over Spectacle
Once Synthesia walked away from entertainment, the company made a decision that defined everything after it.
It didn’t chase open-ended, cinematic AI video. Competitors did that later, and many are still chasing it.
Instead, Synthesia built something closer to a slide-deck editor. Pick an avatar. Type a script. Generate a video in over 140 languages.
The avatars wouldn’t fool a film critic. They didn’t need to. For an HR manager replacing a static PDF, the comparison wasn’t to Hollywood. It was to nothing.
Trust as a Growth Lever, Not a Cost Center
Here’s where I think most operators in regulated-adjacent industries get it wrong.
Synthesia treated moderation, consent, and certification as product features. It became the first AI video company with ISO 42001 certification and holds SOC 2 Type II compliance.
While competitors raced toward open-ended deepfake tools, Synthesia closed doors to misuse and opened doors to banks, healthcare systems, and governments.
In a market full of anxiety about synthetic media, compliance wasn’t friction. It was the unlock.
The Marketing Problem Hiding Inside a Growth Problem
Early traffic looked great. Sign-ups, virality, plenty of buzz.
But the team noticed something I’ve seen in my own campaigns more than once: a wave of “AI tourists.” People who showed up to play with the magic, told their friends, and never came back.
That’s not a product failure. It’s an ai video marketing failure—acquiring curiosity without building habit.
Teaching the Use Case Instead of Selling the Technology
The fix was deceptively simple. Stop marketing the AI. Start marketing the job it does.
Synthesia built templates for specific corporate tasks: onboarding, compliance updates, IT communications. Instead of locking these behind a signup wall, the company published them as search-optimized pages and how-to videos.
Search for “how to make an onboarding video,” and you’d land on a page showing you exactly that, narrated by an avatar.
This is the kind of content strategy I respect most. It doesn’t generate awareness for its own sake. It collapses the distance between discovering a tool and getting value from it.
When the initial hype faded, that content engine kept enterprise buyers flowing in steadily. It also solved the tourist-churn problem, because people weren’t arriving to be amazed. They were arriving to solve a specific task.
What This Means for Operators
Three things stand out to me after two decades of building marketing programs.
The flashiest use case rarely pays the bills. Synthesia’s most profitable customer was a corporate trainer, not a celebrity campaign.
With unfamiliar technology, marketing has to teach, not just announce. Hiding AI behind familiar formats—slide decks, instructional content—does more for adoption than any demo reel.
And compliance, in categories shadowed by public anxiety, can become a genuine competitive moat rather than a checkbox.
The Quiet Discipline Behind the Headline
Synthesia reportedly turned down a multibillion-dollar acquisition approach from Adobe, choosing a strategic investment from Adobe’s venture arm instead.
That’s not a company chasing exits. That’s a company that knows exactly what it built and why it works.
The technology was never the hard part. Teaching an entire market how to use it—quietly, repeatedly, through content built around real jobs to be done—was.
That’s the part most companies skip. It’s also the part that actually compounds.