Most agencies use AI tools. Almost none are AI-native.

You may have noticed I didn’t send a newsletter last week. I was out with a family emergency and I didn’t want to send something that wasn’t meaningful. But that means today I have a lot to say, and I hope you find it valuable.

First….

What really prompted this week’s topic is a LinkedIn reply I received from an agency owner who was very disgruntled about the thought of agencies using AI (they represent themselves as an agency owner). This is where I understand what people are saying - you could tell in his reply he was scared to be out of work because of AI.

But the reality is that it is here to stay, and just like the many other tools and software that have been developed to streamline our work, AI is here to do the same.

You can either fight it, or use it. And understand clients aren’t going to fire you for using it - unless all you deliver is general AI slop. Then yes, that should be what they do.

But that’s not what we encourage. You need to use AI, but give it a brain. Make it understand how you work, your style, train it like you would an employee. Every output has to represent your agency, not just what AI comes up with. You need to give it direction, build it resources, train it, teach it, continue to improve it.

AI isn’t here to do the work for you. It is here to streamline the work, reduce margins, expand bandwidth, and make your team more efficient. It should never replace your strategy, your style, your method. It should complement it, just as an employee would.

And that doesn’t mean it should replace your employees. The focus is on supporting your team. It is true that it maybe replace or reduce some roles, we can be honest about that. But the focus is on building your team up, not tearing them down.

I’d love to hear from you all on your feelings on this. What is making you hesitate to adopt AI, if at all? And on the flip side, if you have adopted it, did you dive in too deep and realize some things just need to be left with humans? Reply here and share where you’re at!

Second….

Something I want to be direct about this week.

Everything we do at Agency Owner Lab is oriented around one outcome: helping agency founders build businesses that don't need them in every room. That's always been the mission. What's changed is how much of that work now runs through AI — and how deliberately we're helping founders get there.

We're focused on helping agencies become AI-native. Not AI-aware, not AI-curious — genuinely restructured so that AI is foundational to how client work gets designed and delivered.

For some founders, that means the done-for-you path: our Agency AI Adoption Engagement, where we map every workflow against the AI-native framework and then build it out — the automations, the Claude systems, the AI agents, the client delivery workflows. We do the build. You run it.

For others — especially at higher revenue levels or who want ongoing strategic partnership alongside the adoption work — that means The Boardroom or the Founder Advisory. Those aren't AI services. They're the human layer: the room of smart peers, the advisor on call, the honest conversations no AI replaces. Founders becoming AI-native still need that. In some ways they need it more.

The two paths aren't competing. They're complementary. Most of the founders we work with closely are using both.

This week we're also releasing the AI Value Audit — a free self-assessment built around the framework in this issue. Map where your agency is today and identify your highest-leverage starting point.

If you want to talk about what the full adoption engagement looks like for your specific business: [BOOK A CALL]

Take two agencies. Both use ChatGPT. Both use Claude. Both have team members with AI subscriptions and some version of "we use AI in our work" in their pitch deck.

One of them will be significantly more profitable, faster, and scalable three years from now. The other will be under pressure — from cheaper competitors, from clients who've figured out how to do more in-house, from a cost structure that hasn't fundamentally changed even though the tools have.

The difference isn't which AI tools they use. It's whether AI is foundational to how they work — or decorative.

That distinction is what separates an agency that uses AI from an AI-native agency.

And the gap between the two is much larger than most founders realize — because it isn't primarily a technology gap. It's an operational design gap. It's about whether AI has been deliberately built into how the business runs, or whether it's been layered on top of workflows that were designed for a pre-AI world.

Let's define the term properly, because it gets used loosely.

An AI-native agency is not one that uses a lot of AI tools.

It's one where AI is designed into the workflow at the point where work is created — not added afterward to make existing workflows faster. The distinction matters because bolting AI onto old processes gives you marginal efficiency gains. Redesigning processes around AI capabilities gives you structural advantages: faster delivery, higher margins, greater consistency, and the ability to scale output without scaling headcount at the same rate.

The three markers of an AI-native agency:

→       AI is involved at the brief stage, not just the production stage — it shapes what gets made, not just how quickly it gets made

→       The team's job is direction and judgment, not execution — humans decide what AI produces, review what it outputs, and own the quality standard

→       Client deliverables are designed to leverage AI production without the client needing to know or care about the mechanics

What most agencies actually look like right now:

AI is used individually by team members to speed up their personal work. There's no shared workflow. No consistent prompt library. No defined standard for what AI-generated work looks like before it goes to clients. Senior people might use it heavily; junior people might not know how. And the founder is still the single biggest bottleneck in quality — because the judgment about what good looks like still lives primarily in their head.

This isn't a criticism. It's where most agencies are in mid-2026. But it means the majority of the available leverage hasn't been touched yet.

An AI-native agency is built in three layers. Each layer has a different function, a different set of AI tools, and a different human role. Most agencies have made progress in one or two. Very few have all three working together.

LAYER 1  Intelligence — Strategy, Research, and Brief Analysis

This is the layer where work starts. Before anything gets created, someone needs to understand the brief, the market, the audience, the competitive landscape, and what a strong strategic direction looks like. In most agencies, this is the founder's job — which is exactly why the founder stays stuck in delivery.

The most important shift in Layer 1: the founder stops being the one who does the thinking and becomes the one who directs, reviews, and approves it. That's a fundamentally different job — and it requires Claude (or equivalent) to have enough context about the client, the brief, and the agency's standards to produce something worth reviewing, not something that needs to be rewritten from scratch.

LAYER 2  Production — Drafts, Designs, Variations, and Deliverables

This is the layer most agencies have partially adopted. AI drafts, AI designs, AI variations. But the way it's usually done — individual team members prompting their own versions with no shared standards — means quality is inconsistent and the human review step is heavier than it needs to be.

The upgrade that makes Layer 2 work properly: a shared Claude Project (or equivalent) with the client's brand voice, past work examples, style preferences, and common feedback patterns loaded in. When that context is shared across the team, AI output is calibrated to the standard from the first draft — not after multiple rounds of feedback.

LAYER 3  Operations — Client Comms, Reporting, Project Management, and Onboarding

This is the layer most agencies haven't touched yet — and it's where the compounding efficiency gains live. The time teams spend on status updates, meeting prep, intake, reporting, and client communication is significant, and almost all of it can be partially or fully automated.

When Layer 3 is working, account managers stop being information routers and start being relationship managers. That shift — from moving information to making judgment calls — is where the real leverage is. And it's what makes a small team capable of managing significantly more client relationships at the same quality level.

THE CEO SHIFT: What the Founder Does When All Three Layers Work

This is the part most agency owners don't see coming — because they're so focused on the individual workflow changes that they miss what happens at the aggregate level.

When all three layers are functioning, the founder's job changes completely.

This isn't theoretical. It's what agency founders who have built all three layers report — consistently. The transition takes 3–6 months of deliberate workflow rebuilding. But the agencies on the other side of it are operating with significantly better margins, more consistent quality, and a founder who has actual strategic capacity for the first time.

That's what AI-native means in practice. Not more tools. A different business.

Before you can build anything, you need to know where you actually are. Most founders overestimate their Layer 2 progress and significantly underestimate the gap in Layers 1 and 3.

Score your agency on each layer — honestly:

Where you have gaps — especially in Layer 1 and Layer 3 — that's where your highest-leverage AI work is. Not another content tool. The workflow redesign underneath.

Layer 1 is the highest-leverage place to start — because it directly replaces the most expensive resource in your agency: the founder's thinking time. Here's a Claude prompt that builds the first version of your Intelligence Layer for a specific client.

Run this once per client, then save the output as a Claude Project for that client. Every new brief goes through this project from that point forward.

Save this as a Claude Project titled "[Client Name] — Intelligence." Add the client's past work, brand guidelines, and any relevant documents to the project. From that point forward, every new brief for that client goes through this project — and the first-pass strategic thinking comes back in minutes, not hours.

Do this for your three largest clients this week. The time investment is about 30 minutes per client. The return is hours of founder thinking time reclaimed, starting immediately.

This week's picks are all directly relevant to building the 3-layer AI workflow — tools agencies are using to actually wire together AI-native operations, not just use AI individually.

Researching accounts. Building lists. Writing sequences.

There's a better use of your team's time.

Apollo is the AI revenue engine that handles the busywork, so you can stay focused on selling.

Plus, everything you need is in one place:

Stop doing busywork and start building pipeline, faster.

With Apollo — the AI revenue engine powering 4M+ users.

Sign up for free

Relevance AI  relevanceai.com

A no-code platform for building AI agents for business workflows — research agents, outreach agents, content agents, and more. For agencies, this is the practical answer to "how do we actually automate Layer 3 without a developer?" You can build an agent that monitors client mentions, drafts the monthly report, or processes inbound new business inquiries — and connect it to your existing tools. Used by agencies at scale to replace repeatable human tasks with AI agents that run around the clock.

Make  make.com

Visual automation builder — often better than Zapier for complex, multi-step AI workflows because it handles branching logic and error handling more cleanly. If you're building the automation triggers for Layer 3 (intake form → brief → Notion → Slack → welcome email), Make is where most agencies are doing that work. Free tier available; paid from $9/month. The learning curve is real but significantly lower than writing code.

Lindy AI  lindy.ai

Builds AI assistants — "Lindies" — for specific repeatable business tasks. Think of it as hiring an AI team member for a defined job: new business research, meeting prep, client follow-up, weekly reporting. Different from a general-purpose AI assistant because each Lindy is trained on a specific workflow with specific context and runs autonomously on a schedule. Relevant for agencies that want Layer 3 automation without building it from scratch in Make or Zapier.

n8n  n8n.io

Open-source workflow automation — self-hosted, which means no per-task fees and full control over your data. For agencies handling sensitive client data who can't route it through third-party automation platforms, n8n is the answer. More technical than Make but significantly more powerful and flexible. If you have someone technical on the team (or a contractor who can set it up), this is the infrastructure-grade option for Layer 3 automation.

Descript  descript.com

AI-powered audio and video editing where you edit the transcript and the video edits itself. For agencies that produce video content, podcast episodes, webinars, or recorded client deliverables — Descript collapses the editing workflow dramatically. Transcription is automatic, filler words can be removed in one click, and AI overdub lets you correct audio without re-recording. Less about the AI-native framework and more about Layer 2 production for any agency with video in their service mix.

📊 BY THE NUMBERS

3–6 months.

That's the typical timeline for an agency to move from "we use some AI tools" to a functioning 3-layer AI workflow — based on what we're seeing across agencies that have done the transition deliberately. Not all at once. Usually Layer 2 first (it's the most visible), then Layer 1 (the highest leverage), then Layer 3 (the most transformational for the founder's time).

The agencies that take longer — 12+ months or stall entirely — are usually the ones that try to tackle all three layers simultaneously without a clear build sequence. The ones that move fastest pick one workflow in one layer, make it AI-native, see the outcome, and use that momentum to build the next.

📣 BEFORE YOU GO

Two things:

First — the AI Value Audit is live. It's the self-assessment tool built around the 3-layer framework from this issue. Map where your agency is, identify your highest-leverage gaps, and get a clear picture of what to build first. Free download: [link]

Second — if you read this issue and thought "I want someone to actually build this for us" — that's exactly what the Agency AI Adoption engagement is. We assess, we design, we build. The entry point is the AI Assessment call. [agencyownerlab.com/ai-adoption]

See you next week.

Work smart. Enjoy life harder.

Founder, Agency Owner Lab

When you're ready, here's how we can work together:

→ The Boardroom — Get in the right room. for 7-8 figure founders. With exclusive partner offers/resources. Spots are opening now! [Apply here]

→ Agency AI Adoption Assessment + Engagement — Custom AI strategy for your agency. Plus option to add 3 months of fractional ops support to make sure adoption sticks. [Apply here]

→ Agency Growth Roadmap — Operational audit + systems strategy. The starting point. [Apply here]

→ Founder Advisory — Your advisor. Your business partner. For the founder who uses AI for strategy but wants a real human to strategize with. Quarterly commitments. [Apply Here]

→ Implementation Sprints — done-for-you systems builds, Standalone or paired with another program. [Book a Systems Audit]