Creator Economy

SWAI's 5,000-Video Playbook: Creator Content Only Pays When Data Rewrites the Brief

SWAI’s Olivia Dumnicka Argues Creator Content Pays Off Only When Each Video Shapes the Next Brief
SWAI’s Olivia Dumnicka Argues Creator Content Pays Off Only When Each Video Shapes the Next Briefclaudid / pixabay

SWAI's Olivia Dumnicka: 5,000+ videos, 1,000+ creators, and a rule that data from every video must rewrite next month's brief — or the budget burns.

By Marcus Bennett

4 min read

At a glance

  • SWAI has produced more than 5,000 videos with over 1,000 creators in Europe and the U.S.
  • Founder Olivia Dumnicka launched SWAI in Berlin in October 2023.
  • Programs need three months before enough data compounds to improve briefs, per Dumnicka.
  • Foodsi has run a nano-creator program (1,000–10,000 followers) with SWAI for almost a year.
  • AI animation that once took a 3D artist weeks can now be done in half an hour, Dumnicka says.

SWAI has produced more than 5,000 videos through a network of more than 1,000 creators in Europe and the U.S., and founder Olivia Dumnicka says that volume only pays off if every video teaches the brand what to brief next. A few creator videos a month used to count as a win, in her account. Today brands need hundreds.

"A lot of brands don't learn from the content that they produce," Dumnicka says. She founded SWAI in Berlin in October 2023 to build creator-powered content engines, running UGC (user-generated content, made by creators for the brand's own channels) and nano-influencer programs that produce social ads, mostly for Instagram and TikTok. Clients include Weleda, Unilever, GymBeam, and Foodsi, in Poland and across Europe.

The model grew out of her influencer work at London tech startups around 2020, when brands paid heavily for single posts from influencers already losing audience trust. "I'd much rather trust my friend's recommendation, or a friend of a friend," she says. "So why don't brands pay normal people who have social media accounts just to recommend something?"

Does follower count predict performance?

No, according to Dumnicka — not when you're buying production rather than distribution. She separates the two purchases: production means content for the brand's channels or ads; distribution means the creator's own audience. Brands routinely ask for a UGC creator, then question why that creator has only 1,000 followers, even though the video was never meant to run on the creator's account.

For distribution, brands judge the wrong signal, she argues: "Oftentimes they are looking more into the general aesthetic of a creator, but not necessarily into their specifically their audience." Cost follows the split. Pricing depends on creator numbers, content volume, usage rights, and whether paid distribution is included — and production costs should stay separate from media spend. Once follower count drops out of the production decision, the same budget buys far more creators.

Why spread one budget across 100 nano-creators?

Because it targets 100 micro-communities and doubles as a test. "If we spread the same budget across 100 nano-influencers, then essentially we are targeting 100 micro communities who might have different problems," Dumnicka says. "It gives us way more data on what works and what doesn't work."

Foodsi, a European competitor to Too Good To Go, has run on that model with SWAI for almost a year, using a few creators monthly in the 1,000-to-10,000 follower range. One, a small business owner, showed the app saving her time on a late lunch; others covered food-waste prevention and healthy cooking, giving one app several distinct use cases.

What should retention data change?

The next brief — and sometimes the casting. The common failure Dumnicka sees is a brand handing an underperforming video's brief to another creator and hoping. What gets missed is hook rate and the second at which viewers drop out, which point to a new hook or a tighter edit. Audience data can change casting too: "Maybe the video was created by a 20-year-old girl, but actually the audience consists of 40-year-olds. Maybe we should change the content creator to be an older content creator, and maybe then the same brief is going to perform better."

Her evidence for the format is client data comparing creator videos against in-house static ads, plus clients who return monthly. Black Friday shows the opposite pattern: rushed clients arrive without briefs asking for 10 videos, and "there is no real time for thinking about the impact of it long term." Her benchmark is patience: "Everything past three months is already when we have enough data to keep improving." SWAI then writes the next month's briefs from winning content and drops the rest.

A trial can also end a program. SWAI planned roughly 100 pieces a month for a vertical wind tunnel attraction, but phones would blow away and glass-shot footage caught reflections or one fixed angle. After a month, both sides agreed the format didn't suit the business.

Where does AI fit?

On the work that costs humans hours. Dumnicka doesn't think AI can replicate humans in spontaneous, camera-heavy footage from creators with strong personalities. Animation is different: a character like Duolingo's owl once took a 3D artist weeks, and "nowadays, you can do it in half an hour." SWAI is testing animated characters for one client, with AI tools building stories from simple directions. She also points to ads styled as 1980s-90s cinema that are cheaper, faster, and perform well; hesitation, in her view, comes from brands who've seen bad AI content.

Her full mix is three sources: creators and influencers, AI, and CGC (customer-generated content — raw videos from people with no ties to content creation). CGC is SWAI's next bet. A UAE coffee shop, for example, might offer 10% off the next visit to customers who scan a QR code and share a photo of their coffee. She expects more results-based pay, citing TikTok Shop, where brands pay only once a sale is made. "This is where the industry is going," Dumnicka says. "There will be no more burned budgets."

Source: netinfluencer.com

Marcus Bennett

Show full bio

Market editor covering money, funding and market shifts around Social platform news and policy at Reach Brief.

More from Marcus Bennett →

Related articles

« Previous articleNext article »