Obvious Playbook, Now Finally Executable: Daily Viral Video at Scale

01 /The Concept

Bluma built a platform that helps brands turn viral videos into a daily content output – at scale.

The platform serves two purposes:

- Publishing what looks like organic user-generated content to TikTok and Instagram accounts.

- Creating video ads for TikTok, Facebook, and Instagram campaigns.

Today's standard workflow for producing this content is fragmented and slow: conceive the video, generate individual scenes across whichever AI video platforms work best for each shot, then assemble them in CapCut or a similar editor.

Bluma compresses that process significantly. Input: a viral video you want to emulate. Output: a finished video about your own product, generated end-to-end by the platform's AI – similar concept, original visuals, ready to publish.

"Similar" here is calibrated carefully. The resemblance to the source video is structural, not literal – same compositional logic, similar character types, same underlying dynamic of what people do and how they do it. You can't be accused of direct copying. But the visual execution is original, which is what distinguishes Bluma's output from the many AI-generated ads that effectively clone each other down to the brand name placement.

The key capability: Bluma's AI identifies the core "hook" of a viral video and converts it into an internal template. That template can then be used to generate videos in the same vein, applied to different products.

The platform ships with a pre-built catalog of these templates, derived from already-successful content, which makes onboarding fast. Clients can also feed in new videos for analysis – generating additional templates that are added to the catalog over time.

Bluma launched recently and is growing at 100% week-over-week, reaching $28,000 in expected monthly recurring revenue after four weeks.

The platform's largest current client uses just 8 templates – but publishes to 16 accounts, twice a day. The scale potential is already visible.

Bluma is currently in Y Combinator.

02 /Why It Matters

The current state of digital marketing has two defining characteristics:

- Short video has become the most effective format. People watch it more frequently and more willingly than any other content type.

- Volume has become the single biggest driver of performance. The stream of video content is enormous – staying visible inside it requires publishing constantly.

But creating videos remains slow and painful. Concept, script, storyboard, footage, asset creation or licensing, editing, performance testing, coordination, publication – and only then do results start to come in.

This process touches several large markets simultaneously:

- The influencer and user-generated content market: $200 billion.

- The video creation and editing platform market: $32 billion.

- The paid advertising market: $180 billion.

In other words, there's as much money in video creation as there is hassle in it.

Some companies have started building in-house "content factories" to put content production on a systematic basis while keeping costs manageable. Targets typically run to 10–15 thousand videos published per month – which includes publishing the same core content reformatted across multiple platforms and accounts.

But even "manageable costs" stay high. Hiring and managing content factory teams is its own operational burden. Many companies lack the resources, organizational will, or expertise to get these factories running.

As a result, some startups are pitching AI-powered content factories – platforms that can produce equivalent volume at lower cost with lower skill requirements.

GetCrux ([related review](/review/kak-sozdat-jeffektivnuju-reklamu)), a Y Combinator graduate, built one of these platforms, describing it as "an AI ad strategist that creates ads that convert." Its logic mirrors Bluma's: the AI monitors what ads are performing well in a client's category, selects the strongest examples, and generates similar videos about the client's specific products.

This "copy what works" principle is also, incidentally, how human content factories achieve their output targets. Their dedicated trend-watchers identify viral videos, break down what makes them work, and commission similar content for their own brands.

Extuitive ([related review](/review/kontent-fabriki-reklamy-tolko-pojavilis-no-ih-uzhe-mozhno-ubit)) raised $20M in its first round in September for a platform tailored to Shopify stores. A store owner provides a link to their store; the AI analyzes the product catalog, proposes advertising formats and visuals for each product, generates corresponding images or videos, and – crucially – tests them against AI-modeled audience segments representing the store's target customers before publishing. Only content that passes testing gets placed.

03 /Opportunities

The overall direction is clear: AI content factories. Without one, effectively promoting products in today's environment is increasingly difficult. Building and running a human-powered content factory, meanwhile, remains too expensive and operationally demanding.

But notice the framing: not "AI platforms" – "AI factories." The difference matters. An AI platform is a tool the client uses themselves. An AI factory is something that delivers outputs for the client at the quantity and quality they need.

GrowthX ([related review](/review/novaja-biznes-model-dlja-bystrogo-i-pribylnogo-rosta)) illustrates why this distinction matters. The startup raised $12M in its first round in May. Its founder had spent years as a CMO running a scalable content production process that grew organic traffic from 1,000 to 30,000 visitors per day in three months – which quadrupled company revenue.

He started teaching this process in paid webinars. Attendees were enthusiastic. Nobody implemented it. They lacked the time, skills, and headcount.

Something clicked: companies didn't need another process description. They needed a contractor who would run the process for them and just deliver the results.

So the founder left his company and built GrowthX as a managed service – same platform, same process, but the client never has to touch it. The results: from $0 in revenue in July 2024 to $700,000 per month by April 2025.

The same dynamic will likely apply to video content at scale. A raw AI platform won't consistently deliver the quality clients actually need – because someone still has to monitor trends, curate inputs, review AI outputs, and iterate. It's less painful than a fully manual content factory, but the friction doesn't disappear. Try to eliminate all of it and quality suffers; keep all of it and clients won't stay.

The implication: the right model may not be purely a software platform, but something closer to an AI-powered agency – outsourcing the content factory entirely, in the style of GrowthX.