Kula, a new UK creator marketing agency, says it can tell brands which creators and content are most likely to meet a campaign's objectives before any budget is committed. That forecasting sits in the Kula Platform, the in-house software the agency has built its whole service around. It launches with seven clients.
The platform combines data, AI and human judgement, according to the company, and keeps learning from live results once a campaign is running. It also handles the work from sourcing creators through to activation, so planning, paid media decisions and delivery sit in one place. Kula's argument is that creative, media and measurement should not be run as separate jobs.
Behind it are Charlie Ashe and Sean Longstaff, who first worked together 16 years ago. Longstaff spent seven years building a social agency and has worked with brands including Domino's and Duke + Dexter. Ashe worked at Ad-Lib, later acquired by Smartly, then ran a consultancy on data, measurement and creative effectiveness, with clients including adidas and Pernod Ricard.
Agencies with their own technology are not new. Billion Dollar Boy used its own platform to track more than 150 brands for the creator share of voice index Hello Partner reported this week, and Socially Powerful introduced its Aria technology in 2024. Kula's emphasis is different, putting the forecast before the spend rather than the report after it.
That is also where the hard part sits. Kantar's research found platform engagement lined up with effectiveness in only a third of cases, as Vera Sidlova told Hello Partner in July. A prediction is only as useful as the outcome it is trained to predict, so brands will want to know whether Kula optimises for views, sales or brand measures.
Kula has not published accuracy figures, named its clients or said how it charges. The fee model matters more than it used to, with brands now funding creators partly from performance budgets, according to Madison & Wall's latest market sizing.
Ashe said AI "is most useful when it helps people make better decisions, not when it replaces judgement". Longstaff said the creative side "can't get lost as brands try to make activity more structured and scalable".
For brands weighing up platform-led agencies, two questions are worth asking up front. How did past forecasts compare with what campaigns actually delivered, and does a brand's campaign data go on to train a model used for other clients?
Building the agency around a forecasting tool is a sensible bet when finance teams want proof before they sign off creator budgets. The quickest way for Kula to win that argument would be to publish how often its forecasts were right.