We covered a lot of ground at the AI Marketing Strategy Summit.

If you’re anything like me, you left both energized and slightly overwhelmed. Between answer engine optimization (AEO), AI-native GTM motions, Cursor rollouts, and hyper-personalized outreach, it can start to feel like there are 800 things you’re “supposed” to be doing with AI right now. And it won’t help when your CEO eventually watches these sessions back and Slacks you about half of these (like mine).

So I treated this summit as my own working session: What would I actually put on my to‑do list as a head of marketing coming out of today? What are the few moves that are both realistic and high‑leverage enough to matter this quarter – not in some abstract three‑year roadmap?

So I’m sharing my to-do list with you, based off of the 8 expert voices we sat down with yesterday.

1. Stand up AI‑powered prospecting and outreach

If there was one pattern across speakers, it’s this: AI‑powered prospecting is quietly becoming table stakes.

Kieran from HubSpot talked about a 30% increase in meetings booked when they layered AI into follow‑ups with inbound leads. Jacob from Grafana walked us through a fully modular agent system that cut manual SDR effort by 95% while improving reply and booking rates.

The common denominator: they are not blasting colder, more generic email. They’re using AI to:

• Read first‑party data (product usage, pages viewed, time on site)
• Infer intent and stage (research vs buy mode)
• Assemble a few truly relevant resources
• Draft an outreach that sounds like a real human who did their homework

My takeaway: if you only do one thing this quarter, make it this. Start narrow. Pick one segment where you already have decent inbound volume and clear behavioral signals (free trials, high‑intent content downloads, product‑qualified leads). Use a workflow tool (Zapier, n8n, Workato) plus your LLM of choice to:

• Generate a short, personalized intro from sales
• Attach 1–3 resources that are clearly tied to what they’ve done
• Log everything so you can see what actually moves the needle
  

This doesn’t replace your best reps; it gives them 10x more at‑bats that don’t feel like spam.

2. Start building your AEO / GEO playbook

This is the one I’ve personally been avoiding the longest, and after today, I’m out of excuses.

Everyone from AI at Work’s Tahnee Perry to Estee Lauder’s Aude Gandon hit the same point: answer engine optimization (AEO, GEO, AIO – pick your acronym) is no longer a thought experiment. We’re already seeing:

• 4-6x higher conversion rates from AI‑chat‑initiated journeys vs classic search
• Early signs of organic traffic erosion in content‑heavy verticals as AI overviews roll out
• AI models leaning heavily on third‑party sources (Reddit, publishers, review sites) to shape their answers

The practical implication: traditional SEO alone won’t defend your discovery surface. You need a parallel AEO playbook that does two things:

1. Earns third‑party authority
   
   • Get your brand and POV into the places models crawl: Reddit threads, niche communities, tier‑one/tier‑two publishers, comparison sites
   • Treat PR and community as inputs to answer coverage, not just brand awareness
2. Restructures your own content for machines and humans
   
   • Build deep, structured pages that answer specific questions in specific ways (not just “ultimate guides”)
   • Use FAQs, schemas, stats, and quotes to make your content easy to cite
   • Map your top problem/solution queries and track whether you appear in AI overviews and chat answers

This doesn’t mean you stop doing SEO. It means you stop doing it in a vacuum. I’m treating AEO as a multi‑quarter program, but we’re starting very simple: audit where we don’t show up today in AI answers for the problems we claim to solve, then work backwards from there.

3. Invest in conversational AI interfaces

The other big unlock is how people actually want to interact now. We heard multiple versions of this:

• DoorDash building one‑to‑one storefronts
• Estee Lauder helping you pick a fragrance or foundation shade through a conversational flow
• Kieran talking about chat as “the new front door” for B2B

Most of our websites are still brochureware with a chatbot bolted onto the corner. The opportunity is to flip that: make conversation the primary interface, not the accessory.

That doesn’t mean throwing a generic LLM on your homepage and hoping for the best. It means:

• Training on your product, docs, and real sales conversations
• Designing flows around actual jobs to be done (“Help me pick the right plan,” “Compare X and Y, “Show me how to do Z with your product”)
• Deciding explicitly what the bot should own vs when to hand off to a human

At Section, this is forcing us to rethink what a “page” even is. A lot of what used to live in static comparison tables or pricing grids is better handled as an interactive Q&A. The long‑term vision is: whether someone lands via search, AEO, direct, or email, they should be able to talk to us, get something useful fast, and move forward without hunting through nav menus.

If you’re early here, don’t try to boil the ocean. Pick one high‑value flow (pricing, plan selection, onboarding) and build a genuinely good conversational experience for that, end‑to‑end.

4. Give your marketing team Cursor (and expect them to use it)

This might have been my favorite uncomfortable moment of the day.

Dan from Zapier didn’t just say “we encourage AI use.” He said: we mandated a 30‑day Cursor rollout for every marketer, tracked who was building what, and made it clear this wasn’t optional. And then they backed it up with support: workshops, pairings, and a shared expectation that marketing is now a builder function.

The reason this matters: there’s a big difference between “I use ChatGPT to rewrite intros” and

I can work with agents, APIs, and data to build small systems that change how my job works.

I don’t think every marketer needs to become a full‑stack engineer. But I do think the bar for AI proficiency is rising fast, and tools like Cursor (or other AI‑native dev environments) are the bridge between copy‑pasting prompts in a browser and actually automating workflows.

Practically, my plan looks like:

• Standardize on one dev‑adjacent tool (for us, likely Cursor) that everyone can access
• Run a 30‑day “build something real” sprint where every marketer ships at least one live workflow or internal tool
• Track it: who set it up, what problem it solved, what broke, what we learned
• Make this part of performance and career conversations, not a side hobby

The goal isn’t to turn your creatives into Python wizards. It’s to make sure your best thinkers are no longer blocked by “someone technical” every time they see a system problem worth fixing.

Your mandate

You do not need to do all of this by next Tuesday. But you also can’t wait a year and hope this all blows over.

Pick one of these four, go deep for 60-90 days, and make it impossible to go back to the old way of working in that area.

Then move to the next one.

If you’re feeling behind, you’re not alone. I run marketing at an AI company, sat through this entire summit, and still had the ‘I thought I was ahead and now I’m back at the beginning’ moment when Dan started talking about universal Cursor adoption.

The good news is the gap isn’t between you and some mythical “AI native” competitor. The real gap is between the slides you’ve seen about AI and the first two or three concrete systems you build that actually change how your team works.

That’s the gap I’m focused on closing this quarter.