Content Alchemy · Service 03
Technical SEO Audit. Hyper-realistic synthetic personalities, multilingual and always-on. Crawlability, indexability, Core Web Vitals, schema, internal architecture — examined with surgical rigour, prioritised by commercial impact.
A founder has 24 hours in a day.
A spokesperson can fly to one country at a time.
A model needs to be in front of a camera to deliver a line.
An AI avatar has none of those constraints.
One identity. Five languages. Every channel. Forever.
The difference between a good AI avatar and an uncanny one isn't the rendering technology — it's the character work behind the face. We don't start with the model. We start with the personality: what they believe, how they speak, what makes them recognisably themselves across thousands of pieces of content. The rendering is the easy part.
For B2C brands we typically build a fictional spokesperson — designed from scratch with a defined backstory, look, voice and personality that matches the brand. For B2B and founder-led companies we build a founder avatar — using real photos, voice recordings and behavioural reference to produce a synthetic version of the actual founder, capable of appearing in five languages without translation costs or scheduling.
Every audit covers four pillars. Each pillar is a deep dive — not a checklist tick.
Before any face is rendered, we define who the avatar is. Backstory, worldview, voice principles, behavioural patterns, what they would and wouldn't say. Six pages of character work that everything else builds from.
For founder avatars: MetaHuman Identity capture from real photographs and reference video. For fictional spokespeople: original character design built in MetaHuman Creator + Character Creator 4. Either way: visually indistinguishable from a human in 4K video.
For founder avatars: 15-minute voice training in their native language, then ElevenLabs cross-language voice extension to deliver in 4 additional languages. For fictional avatars: original voice synthesis tuned to match the character's personality.
Once built, the avatar is operated as a content asset: scripted, rendered, lip-synced (AccuLips), motion-captured (iClone), composited (Unreal Engine), and shipped across channels. Typical output: 12× the volume of an equivalent human-only production line.
2-3 weeks of structured workshops to define who the avatar is, what role they play in the brand, what content they'll deliver, and what we deliberately do NOT want them to do. The brief is the foundation.
4-6 weeks for visual identity (MetaHuman + Character Creator + Unreal Engine) and voice (ElevenLabs + AccuLips). Multiple iteration rounds. We don't move forward until the avatar is indistinguishable from a real person in motion video.
2 weeks of producing the first batch of content (typically 20-40 pieces across formats and languages). The pilot proves out the production pipeline and catches every edge case before scale.
Monthly retainer thereafter. Content briefs in, finished video out. Typically 30-100 pieces per month per avatar across multiple languages and channels. The avatar becomes a permanent operational asset.
A real spokesperson costs €5,000-15,000 per shoot day, requires scheduling weeks in advance, can deliver in one or two languages, and creates a single batch of content per session. An AI avatar — once built — produces unlimited content, in any language, on any timeline, at the marginal cost of compute.
The break-even point against a traditional spokesperson programme is typically 3-4 months. After that, the avatar becomes a compounding content asset — getting more refined, more on-brand, and more efficient with every production cycle. Five years from now, the brands that started in 2026 will have an unfair operational advantage.
Infinity8 Cosmetics — a B2B private-label nail and cosmetics brand — needed a recognisable face to anchor their content programme to professional nail technicians. We designed Selene: 26-year-old, Franco-Italian aesthetic, warm chestnut-auburn hair, Bardot rose-beige lip, hyper-real. She speaks four languages natively, demonstrates products on her own (synthetic) hands, and delivers 12× the content output of an equivalent human-only programme.
We needed Selene to be more than a face — she had to be a personality professional nail technicians would recognise as one of their own. The character work is what made it work. By month four she was generating more inbound enquiries than our entire previous content programme combined. The economics aren't comparable to anything in traditional content production.
A technical audit is most powerful when followed by these complementary services.
For brands that need traditional human-led video alongside avatar content — or for the launch campaigns that introduce the avatar to the world. Our video team works directly with the avatar pipeline.
Explore →An avatar without a content strategy is just an expensive face. The strategy defines what the avatar will say, where they'll appear, and how the content compounds over time. Strategy first; production second.
Explore →For pre-launch brands, the avatar is part of the brand expression — and needs to come downstream of strategic positioning. We typically run brand strategy and avatar development as paired engagements.
Explore →The future of brand presence
Book a 30-minute discovery call. We'll discuss whether a fictional spokesperson or a founder avatar is the right approach for your brand, walk through the production pipeline, and quote a possible engagement with no obligation.