AI Agents at Agency Jellyfish Cut Campaign Launch Times by 65%

Brandtech Group-owned agency Jellyfish is using AI agents to carry out media buying tasks across Google, Meta, and Amazon, from automating campaign setups to optimizing performance, significantly reducing the need for large teams and slashing campaign launch time by 65%. The AI agent—which carries out and completes tasks on behalf of people—mimics the decision-making process of a junior-level media buyer. It handles an array of tasks, like organizing campaign data, setting up reporting dashboards, and adjusting budgets daily to avoid under or overspending—duties that once took weeks and demanded continuous human oversight. “Clients launching a new campaign today can have a reporting dashboard set up the same day [with AI agents] instead of waiting up to 40 days,” said Jeff Matisoff, partner at Jellyfish. The agency manages over $2 billion in media budgets. With the introduction of AI agents, what was once a time-consuming process, especially for performance marketers, is now completed in a matter of hours. This shift is changing ad-buying operations, enabling brands to quickly pivot and fine-tune strategies based on real-time data. In line with this growing trend, The Brandtech Group has made significant strides in AI investment, including acquiring the generative AI-as-a-service platform, Pencil. CEO David Jones told Bloomberg last year that Pencil has managed over $2 billion-dollar media spend across 5,000 brands since 2018 and has created millions of generative AI-driven ads. Jellyfish’s investments in AI agents are already paying off. The AI agents have reduced infrastructure expenses by an average of 22% and increased campaign performance by 30% across several clients, according to the agency. “Our technology updates every 15 minutes, catching mistakes and fixing them in real-time,” Matisoff said. Digital employees on the rise The agency began integrating AI agents about eight months ago, treating them as digital employees. Just as new hires at Jellyfish are trained to log into platforms, set up audiences, and ensure brand safety, the AI agents learn these tasks through APIs and automated frameworks. Once a client approves a plan—outlined in slides and spreadsheets—the agency's human team traditionally logs into platforms like Meta or Google to set up campaign tasks such as insertion orders (IO), keywords, assets, and targeting. Now, the AI agents perform these same tasks automatically, carrying out media buying across Meta, Google, and Amazon—a majority of ad spend goes across these platforms, according to Matisoff. “The agents are replacing the work done by the humans,” Matisoff said. “In-house marketers that don't want to have dozens of junior employees on their books now have these digital employees deployed to do exactly that same thing. That takes teams of thirty or forty down to a team of four.” Human approval is still needed Each brand account at Jellyfish is paired with an AI agent, operating within pre-defined guidelines specific to a brand’s KPIs and budget allocations. These agents execute tasks within those parameters, while larger decisions—like shifting platform budgets or testing new audiences—still require human approval. For example, if a client approves budget shifts between Amazon and Google Ads based on a minimum return on ad spend, the AI agent can make those adjustments autonomously without the need for human intervention. “It's broadening our ability to solve problems that are far more integrated from the start,” said Natasha Wallace, managing director of Jellyfish in the Netherlands. Enter the multi-modal strategist According to Wallace, the adoption of gen AI and the automation of marketing tasks has “changed and migrated the types of skills we want to have in the people working in our teams.” Among these new skill sets is a role that has emerged within the agency: a multi-modal strategist. This was an internal hire and the person is tasked with synthesizing insights and transforming them into actionable strategies across retail, social, creative, and paid media. “Our approach involves both up-skilling existing talent and nurturing promising individuals to operate effectively across multiple domains,” Matisoff added.

Artificial Intelligence

AI Agents at Agency Jellyfish Cut Campaign Launch Times by 65%

AI is being used to automate campaign setup, optimize performance, cutting weeks-long processes down to hours

UNLOCK FULL ACCESS

Trishla Ostwal

Media

TikTok Expands Its Off-Platform Ad Network to US Advertisers

Piers Morgan's Uncensored Launches Paid Tier to Own Its Audience Beyond YouTube

OpenWeb Enters Insolvency After Dramatic Fall From $1.5 Billion Valuation

Condé Nast CEO Roger Lynch Departs for Mattel

OpenAI Takes On Meta's Muse With Dots, an Always-On AI Agent

OpenAI Is Piloting Brand Suitability Controls With DoubleVerify and IAS

Paradium.AI, Formerly The Arena Group, to Sell 49% Stake to Roundtable—If It Can Raise $89M

Global Citizen Taps Influential As Its Official Creator Partner

The Trade Desk, Worried Stock Slide Could 'Substantially Harm' Retention, Moves to Reprice Employee Options