Agents in Action 2026: Key takeaways you don't want to miss

Agents in Action (the AI virtual event built specifically for marketing and digital teams) made its 2026 debut, and it absolutely delivered.

We brought together a powerhouse lineup of speakers, including Nathaniel Whittemore (host of The AI Daily Brief), Daniel Hulme (Chief AI Officer at WPP), Bernard Huang (Founder of Clearscope), John Habib (Director of Content Strategy at Diligent), and Greg Kihlström (Principal at The Agile Brand).

Alongside our own AI enthusiasts at Optimizely — ft. Julia Maguire, Ali Hart, Mark Wakelin, and Leah Messenger — they tackled everything from where humans provide the most value in an agent-driven world, to how to show up in AI-powered search, to the marketing use cases your competitors haven't caught onto yet.

Missed it? Don't worry, it's all available on demand. But first, here's your sneak peek at the key takeaways.

1) AI's making content infinite, so double down on your human-only traits

We've reached an inflection point. AI has pushed us past the stage where creating content is the bottleneck. Our guest speaker, Nathaniel Whittemore (host of AI Daily Brief) put it plainly: production is now bountiful, if not infinite. Think beyond simple A/B tests; we're talking a "Doctor Strange approach" to marketing, simulating millions of scenarios to find that sweet spot.

So, if AI can crank out endless content, what's your superpower? Nathaniel argues it's all about the uniquely human touches that AI simply can't fake. The best definition of brand, he reminded us, is the emotional aftertaste of an experience — and AI doesn't know real-time cultural vibes. Your intuition, empathy, and instinct for what's going to resonate right now? That's gold.

But here's the kicker: volume alone isn't enough. Spinning up a huge amount of content is different from getting it into a pipeline where it can actually flow. This is where smart systems and vertical AI come in; tapping into deep domain expertise and real endpoint data from customer interactions to build the pipelines that make AI genuinely work for you. It's about redesigning how you operate, not just adding tools.

Watch Nathaniel Whittemore's full keynote on demand to learn how to amplify your human edge and build systems built to thrive.

2) Most AI failures start with organizational problems, not technical ones

Daniel Hulme (Chief AI Officer, WPP) and Julia Maguire joined forces to explore the evolving relationship between humans and agents in marketing — and the results were eye-opening.

Your newest "customers" won't always be human. AI agents acting on behalf of people are already entering the picture, which means brands need to become legible to agents: clear terms, structured data, verifiable claims. And in a world full of synthetic content, trust becomes the most valuable currency a brand can hold. Provenance, authenticity cues, and responsible generation policies are no longer optional — they're core to brand strategy.

The other big reality check? Most AI failures aren't technical. Bad data, unclear ownership, and weak goal-setting are the typical failure points. Fix the foundations before you layer on more tools.

Catch the full session on demand.

3) Centralize, templatize, and scale

John Habib (Director of Content Strategy, Diligent) shared how his team went from AI curiosity to a genuinely agentic marketing engine — and the approach is replicable.

The starting point? Document your processes. Get honest about where your gaps and friction points are, because that's exactly where AI can plug in and make a difference. Then standardize everything: templatize the inputs for common projects — webinars, blog posts, web pages, social content — so agents can quickly and efficiently produce the right outputs without reinventing the wheel each time.

And critically: don't set it and forget it. AI is constantly evolving, and even the best agents need to be regularly refined to continue delivering the quality and consistency you need.

Watch the full Diligent session on demand to see what's possible.

4) Don't stress — AEO is actually very similar to SEO

Traffic is down. You're not showing up like you used to. Sound familiar? Bernard Huang (Founder, Clearscope) broke down the new reality of AI-driven search — and made it a lot less intimidating.

The future of search is conversational. Users increasingly prefer chatbot experiences over the traditional "10 blue links," which means the game has shifted. But here's the reassuring news: Agent Engine Optimization (AEO) is actually very similar to SEO. The main difference is how you get rewarded — visibility rather than direct traffic clicks.

And the fastest path to AI visibility? It's simpler than you think: rank higher on traditional search engines like Google or Bing. The fastest way to influence what AI says about you is to be highly ranked where AI goes to learn.

Get the full AEO playbook by watching the session on demand.

5) When it comes to finding use cases, start with pain points (not possibilities)

Julia Maguire's session cut through one of the most common traps in AI adoption: chasing possibility rather than solving real problems.

AI pilots fail when they're applied to vague or broken workflows. A well-defined problem is the strongest predictor of measurable impact. The highest-value AI use cases consistently come from understanding where work is slow, manual, fragile, or overly dependent on specific individuals; the bottlenecks hiding in plain sight across your marketing lifecycle.

A shared framework for surfacing these opportunities makes all the difference, helping teams align on priorities, pressure-test ideas, and move from one-off pilots to scalable agent use cases.

Watch the full session on demand to get the framework.

6) Design for experience value, not just your own team efficiency

Greg Kihlström (Principal, The Agile Brand) and Ali Hart challenged the room to think bigger about what AI is actually for.

Yes, AI can save time and reduce costs. But it delivers its biggest impact when it genuinely improves customer and employee experiences — not just operational metrics. Speed becomes a competitive advantage only when the messages and experiences maintain the nuance and intent of your brand voice. At scale, that requires automated brand governance.

The other big shift? Moving from point solutions to platform thinking. Building a truly connected, AI-first ecosystem means thinking in terms of a comprehensive customer platform — not channel-by-channel or tool-by-tool.

Explore the full session on demand.

7) Make ownership and responsibility explicit if you want to scale your AI program with confidence

Mark Wakelin closed things out with the governance conversation that too many teams avoid — and the one that determines whether AI actually scales.

The first lesson: choose a governance model that matches your AI readiness. Teams stall when AI is treated as a one-off tool rather than a program. The right model enables speed while protecting brand, data, and compliance.

The second: make ownership explicit. AI only scales when roles are clearly defined — who sets guardrails, who builds and maintains agents, who owns quality and risk. Ambiguity is where programs stall.

And finally: measure outcomes, not just activity. To unlock the next wave of AI value, you need proof — track productivity, performance, and business impact so you can prioritize what to invest in next.

Watch Mark's full session on-demand, and take home the comprehensive AI Marketing Playbook.

8) The Agent Lab: Where you can see agents... in action

To cap off the event, attendees got a front-row seat to real agents built by real marketing teams in The Agent Lab. Here's what was showcased:

• Virtual Focus Group Agent (built by Netcel): Get instant, persona-driven feedback to create more effective and personalized content at scale, without the cost of traditional user research.
• Competitive Benchmark Agent (built by Robinson Club GmbH, part of TUI Group): Automate days of competitive analysis into minutes, with strategic recommendations and on-brand content to boost marketing velocity.
• Content Publish Update (CPU) Agent (built by Diligent): Save hours each week by automating the manual consolidation of content data into clean, structured reports for internal enablement.
• Webinar Campaign Builder Agent (built by Diligent): Automate the entire creation of a webinar campaign, from task setup and due date calculation to generating on-brand draft assets.

That's just a taste of what Agents in Action had to offer. Every session is now available on demand — packed with actionable frameworks, real-world examples, and the kind of practical insight that actually moves the needle.

Watch all sessions on demand →