When your core technology carries a reputation for harm, aim it at a boring high-value job and build consent, safety, and moderation directly into the product so trust becomes your competitive advantage.
Synthesia
Synthesia is a London-based AI video platform, founded in 2017, that turns typed text into presenter-led video using synthetic avatars and voices, aimed squarely at enterprise training, onboarding, and internal comms. Its through-line is discipline: while others chased consumer deepfake spectacle, Synthesia pointed generative video at a boring, high-value corporate job, built consent and moderation into the product as a trust moat, and grew from a research demo into a company crossing $100M ARR in 2025 by making that job self-serve.
Differentiation: consent-gated avatars in a market defined by deepfake fear
The problem. Synthesia's core technology, cloning a real person's face and voice from text, is the same capability that powers malicious deepfakes, which made the entire category radioactive to the enterprise buyers Synthesia wanted (training and comms teams inside regulated Fortune 100 firms). A platform indistinguishable from a deepfake tool could never win a procurement review.
The approach. Synthesia built trust and safety into the product rather than bolting it on, codifying an ethics framework it calls the "Three Cs": Consent (an avatar can only be created with the person's explicit verified consent), Collaboration (with regulators and industry bodies), and Control (customers decide who uses their avatar, when, and how). It moderates 100% of content through automated and manual review and became a launch partner of the Partnership on AI's Responsible Practices for Synthetic Media framework.
How it solved it. The guardrails became a differentiator instead of a constraint. In a red-team exercise run with NIST and Humane Intelligence, testers made over 40 attempts to create non-consensual avatars and Synthesia's systems detected and blocked all of them, giving enterprise buyers a defensible answer to the deepfake objection and helping the platform reach roughly 70% of the FTSE 100 and over 90% of the Fortune 100.
Category Creation: making "AI video" a boring enterprise line item, not a party trick
The problem. In 2017 there was no "AI video" category for businesses to buy. Generative video existed mostly as viral novelty and research demos (Synthesia's own 2018 BBC segment had newsreader Matthew Amroliwala "speak" Spanish, Mandarin, and Hindi), which is memorable but not a budget line any L&D or comms team knows how to purchase.
The approach. Synthesia named and framed a concrete job to be done: making corporate video without cameras, studios, actors, or editing. CEO Victor Riparbelli anchored the pitch on driving "the cost of content creation down to zero," reframing the technology from spectacle into a mundane productivity tool for training, onboarding, product demos, and internal communications.
How it solved it. By defining the category around a repeatable enterprise workflow rather than novelty, Synthesia became the default answer, serving over 60,000 customers and growing revenue from about $42.8M in 2023 to $62M in 2024 and past $100M ARR in 2025, and closing a $200M Series E in January 2026 at a $4B valuation with backers including Google Ventures and NVIDIA's NVentures.
GTM: pivoting from bespoke agency projects to self-serve enterprise SaaS
The problem. Synthesia's early go-to-market was effectively an agency and API model, producing custom AI videos as one-off projects, which is high-touch, hard to scale, and does not compound into recurring revenue. A demo-famous technology was not yet a repeatable business.
The approach. The company pivoted to a self-serve, browser-based SaaS product (Synthesia STUDIO) that lets any employee type a script and generate a professional avatar-led video in minutes, then layered enterprise motion (security, multi-seat, localization into 140+ languages) on top of that product-led base to land and expand inside large accounts.
How it solved it. The self-serve product turned a services business into a scalable SaaS engine: Synthesia now derives roughly 70% of revenue from enterprise deals while serving over 60,000 customers, reaching more than 90% of the Fortune 100 and 95% of the DAX 40, and the model let the company scale ARR past $100M in 2025 without the linear cost of custom production.