Why Brands Are Pulling Ad Production In-House With AI

The agency business spent a year telling clients that AI would make their creative cheaper. It did. The clients noticed they could run the software themselves.

Twenty-Four Days to Two Hours

Kimberly-Clark built an AI content platform out of its India operation that cut the time to produce a piece of marketing content from 24 days to two hours, according to Marketing Tech News. Deena Dayalan, who heads Kimberly-Clark India, said the same tool handles influencer selection and campaign localization across markets.

That is not an isolated experiment. Catalyst Brands, the parent of J.C. Penney, is testing computer-generated product images and video for online listings, with managing director Nihar Nidhi telling Reuters the approach could reduce the need to ship inventory around the world for photo shoots. At Target, the Roundel ad arm has copywriters using AI to produce ads faster, per Target India president Andrea Zimmerman.

What the Brands Are Actually Internalizing

The work moving in-house is the production-heavy middle of the funnel: image and video generation, copy variation, localization into dozens of markets, influencer shortlisting, and the testing loop that picks winners. These are exactly the tasks agencies used to bill for by the hour.

A survey by the World Federation of Advertisers and The Observatory International found that 66% of major multinational brands already run an in-house agency of some kind. AI lowers the cost of standing one up, because the headcount that used to be required to operate a creative pipeline is now partly a software license. The marginal cost of one more ad variant falls toward zero, and the work that justified an external retainer goes with it.

The Agency Value Proposition Is Being Repriced

Gartner analyst Jay Wilson put the squeeze plainly: agencies can no longer rely on production capacity as their advantage, so strategy and specialized expertise become the differentiator. That is a real business, but it is a smaller and more senior one than the staffing-heavy model most agencies still run.

The uncomfortable part for agencies is that the brands pulling work in-house are not the small ones. Kimberly-Clark, Target, and J.C. Penney are exactly the accounts that paid for the big retainers. When those clients internalize production, the agency loses its most profitable, most repeatable revenue and keeps the lumpy strategic work that is harder to staff and harder to scale.

Why Regulated Firms Cannot Just Copy This

Here is where the trend hits a wall. A consumer-goods brand can let an in-house model generate a thousand product shots and ship the good ones. A wealth manager, an insurer, or a bank cannot. Every claim in a financial ad is a potential regulatory event, and an unsupervised model that writes a performance figure or an implied guarantee creates exposure the moment it publishes.

So the firms that would benefit most from in-house AI production are the ones least able to use the off-the-shelf consumer tools. They need the speed of Kimberly-Clark's pipeline with a compliance constraint built into every generation. That combination does not come out of a generic image model pointed at a product catalog.

The In-House Engine for Compliance-Bound Marketing

This gap is the reason LeadLord exists. The product grew out of a wealth firm that spent five months and $100,000 with an outside agency and shipped zero campaigns, because the agency could not meet the firm's compliance requirements. LeadLord is the in-house engine that firm needed: copy, creative, image generation, hosted landing pages, and multi-platform A/B testing on Meta, Google, and LinkedIn, with compliance rules living inside the draft.

The Kimberly-Clark story is what that looks like in consumer goods, where the only cost of a bad asset is a wasted impression. LeadLord applies the same in-house, full-stack model to industries where a bad asset is a filing. The pitch is the same disintermediation the big brands are running, aimed at the firms that were locked out of it because their content has to clear a regulator before it ships.

What to Watch

The open question is how fast the regulated end of the market follows the consumer end. Consumer brands have a two-year head start because their downside is low. Finance, insurance, and healthcare move when the tooling can prove it controlled the output, not just generated it. Watch for the first wave of in-house AI marketing teams inside regulated firms, and watch whether they build the compliance layer themselves or buy it. The firms that figure out controlled in-house generation first will book the same cost advantage Kimberly-Clark just did, in a market where the agencies were never able to serve them well anyway.