Most Agencies Are Failing at AI — Because They Have No System

Hey everybody, my name is Macgill. I'm co-founder of Rize and I'm really excited today to sit down with Leah Leaves. Leah is an expert in the agency space and we're going to be chatting today about AI and how it's transforming agencies. Leah, do you want to give a little introduction?

Happy to, and happy to be here. This is a fun topic for me because I've been in the agency space for a long time, specifically in agency operations. I love this space because it's all about supporting creative, incredible visionaries — marketing strategists and their teams — by creating the right infrastructure and support mechanisms to amplify their value to customers. My company, Alderaan Operations Solutions, places fractional operators and project managers to help with that from the inside out. We love being the order-out-of-chaos makers.

I can't imagine being in a better position to see how AI is actually changing agencies. At a really high level, what's the top one, two, three ways you're seeing AI immediately change how agencies operate today?

I'll split it into a binary of what's really good and what's really bad. The good is that AI is doing what good operations support does — amplifying what agencies are already great at. Their secret sauce, the strategic value of marketing and advertising partnership. When agencies embrace AI in creative ways, it gets them past grunt work and administrative functions and lets them deliver the top-line, value-add work. That's AI 101 for agencies.

The bad side is companies pushing back on agencies because they're saying, 'I'm using AI too. If I can put my marketing skills into Claude or ChatGPT and just have it produce AI slop that's good enough, what does the agency do differently?' If the agency is still focused on execution-only — a long list of deliverables, ten blog posts a month, a laundry list — there's no differentiator, no strategic or advisory layer. That's where agencies are getting squeezed. Either you lean in, empower your team, become more creative and strategically valuable, or you get squeezed out.

What strategies would you recommend for agencies in that situation — to make sure they're focusing on amplifying and moving away from being just execution?

A practical example we've seen across agencies is monthly client reports. A year or two ago, agencies started using AI to help generate these reports — that's still in the execution layer. The agencies flipping to the strategic layer are mining those reports for gold nuggets, then bringing proactive recommendations to client calls. That's account management 101 — proposing new ideas, surfacing channels the client hasn't considered, being proactive.

Some agencies just use AI for report generation — great, you've taken admin off your plate, but if you're not actually reviewing the reports or cross-referencing them with AI meeting notes you've been collecting for years, you're missing the analysis, interpretation, and strategy that's the real value-add. A handful of agencies have run with this tenfold and built their own internal products around the data — pulling reports, campaign metrics, and meeting notes into benchmarking products that take the service even further. The ones doing well take the strategic insights into client calls and actually do something with that data — they don't just send it in an email.

We feel this same pressure at Rize. We want to use AI to streamline our processes, but one thing we run into is: where do we start? I've spoken to agency owners who feel the same way. Do you have concrete recommendations? When you start working with a new agency, what are some ways they can start building that AI muscle?

We hear that often — agencies don't know where to start, and they feel that anytime they do start, they're behind the next day because a new feature drops or a new platform launches. We look for two things: ownership and a strategic plan.

First, ownership. One of the biggest gaps over the last few years is that only the owner is owning AI strategy. They're the ones pushing it forward, getting excited, tinkering on weekends. Some team members are playing with things, but it's scattershot — spaghetti at the wall, no structure. Whoever owns it — the founder, the CEO, or someone you designate — needs a real seat on the organizational chart with clear role responsibilities, not just an add-on to their existing title.

Second, the strategic plan. I'm not talking about a three-year plan — things are moving too fast for that. But you need some concept of where you're heading. We look at it in three layers. The first is the 30,000-foot view: what is our AI vision, our ethics, our guardrails? How are we talking about AI in our company culture? That informs recruiting, HR, and performance management.

The second layer is systems. What does this practically look like? Can we get everyone onto one platform? I know agencies with team members using 20 different tools, all tinkering, debating which is better, while 80% of the team just uses ChatGPT Teams. Big-picture culture informs the systems — pick a primary LLM, do a company Teams account, build infrastructure.

The third layer is initiatives — what projects are we actually going to focus on? Maybe you have a form your team submits ideas through, and the AI owner takes those, works with leadership, and crafts what to focus on next. Maybe it's client reporting, with clear definitions of what good looks like. Now it's not esoteric — it's tied to actual company goals.

After working with agencies of all sizes — from half a million in revenue to a hundred million — most are still dealing with client retention and client churn. That's almost always a goal. So a low-hanging first initiative is: how do we layer AI in to support increasing client retention? That ties everything together — ethics, systems, policies — into a 30, 60, 90 day sprint. Run a couple of experiments, measure them quantitatively and qualitatively, and pair the AI owner with the AI plan. Now you have a starting point — it doesn't feel wild west anymore.

Having someone designated with that role makes total sense — it's the only way I've ever seen anything actually get done. Is this a role agencies are hiring for, or are COOs taking it on? What titles are you seeing?

In 2026 we're seeing it as an actual separated seat on the org chart. In 2023, 2024, 2025 it was usually appended to someone else — typically the owner or COO. A lot of agencies created mini focus groups or AI councils with representatives from multiple departments, but where that fell down was no single owner leading the council, so good ideas didn't move forward.

Coming into this year we're seeing real titles: AI Strategy Lead, AI Director, and sometimes AI Operator or AI Operations Lead. The difference depends on whether the agency expects a unicorn — strategic vision, planning, and execution all in one. That's like expecting one person to be both marketing strategist and project manager. They're different seats for different reasons. We're seeing more agencies split it: AI Strategist plus AI Operator who thinks in systems and process and can execute.

One area agencies are expecting but haven't put in writing is AI literacy. They're expecting basic AI fluency in everyone, but no one is adding it to recruiting expectations or quarterly performance check-ins. They assume team members are doing things with AI, but no one writes down what's actually expected. So while titles like AI Strategic Lead and AI Engineer get figured out, the bigger caution is: yes, you need an owner and a leader, but it should not be 100% on their shoulders. Everyone needs to be learning, contributing, and engaged — just like every other initiative. AI literacy and fluency will be a real focus for the rest of 2026.

We've actually been seeing a use case at Rize where AI strategists or AI implementation officers use Rize as the metric to see if their AI strategy is working. They have Rize running for the team measuring tasks, then they can actually see whether task velocity increases or decreases — and feed that back through AI for analysis. Interesting to hear that's becoming common.

I love that. It's so critical for agencies because the bigger question is: what's the ROI on all this? How do we bill for it? How do we change pricing? Whether it's an AI agent or a human on your org chart, you still need to track it — that data feeds the strategic decisions. I love that you already have that and are playing with it.

AI is changing billing, profitability, and margins for agencies. What pricing models do you think will win out? Where do you see this going?

My thoughts haven't really changed with AI — I was always biased toward value-based pricing over time and materials. There's always been tension between execution and strategic layers, and when an agency is too heavy on execution, it's easy for the customer to comparison-shop on price because there's no strategic differentiation. AI has just added fuel to the fire. The customer comes back saying, 'I can produce the same thing with ChatGPT.' They can't, really — they don't have the nuance, strategy, or context — but on paper they see an output and assume it's equivalent.

If the agency team isn't trained or doesn't have the reps to explain the strategic value — the years of collective experience — then you're stuck in price-comparison shopping instead of being a strategic value-added partner. We're still seeing agencies with strong five- or ten-year client relationships getting questioned now: 'I assume you're using AI, so you must be more efficient. What more are you going to give us?' Sometimes clients want more deliverables for the same price; sometimes they want the same outcome cheaper. It's a delicate balance.

Long-term, value-based pricing has always been the space to be in, and now even more so. If you haven't leaned into it or evaluated what it looks like for your agency, this is the time. Otherwise you're going to be in the bucket of agencies getting squeezed out.

How are you recommending agencies think about margins? Are the agencies that are well along the AI adoption curve actually tracking AI output and AI engagement from teams? Are they tracking it in terms of cost and margin?

Short answer: no. Agency owners still don't value all of those operational metrics because they don't always see the one-to-one impact. What I do see them tracking is top-line revenue. The more mature ones track gross and net profit margin. How they incorporate AI into top-line is by adding services — AI advisory, AI strategy input for clients, sometimes AI training. But when we work with them we have to remind them: track the costs too. That's where you get into gross and net profit margin, team engagement, and project velocity. Most agencies aren't yet incorporating AI into their full 360 view of finances. They're just asking, 'Is this helping us make more money?' and they don't have the systems — like Rize — to properly track it.

We're working on agent session tracking inside Rize — so you can tie agent sessions to client work and see profitability and ROI from both human delivery cost and agent cost. It's so new, but a lot to build.

You're absolutely on the right track. We've been beating the drum of having a bionic org chart for a few years. Anytime you're adding something or someone to a function, identify it on the org chart — whether it's a tool, a platform, or a person doing the job. Agency owners almost always tell us about delivery and service when we ask about their org chart — strategists, specialists, VAs. Then we ask: what about bookkeeping, taxes, HR, recruitment, marketing, sales? Probably ten seats, all with your name on them.

There are still ten functions. Now that AI is producing some of them, it's still a seat on the org chart. It's a custom GPT that does all of my sales admin? Great — give it a name, put it on the org chart, label it, identify it. That supports pulling in a tool like Rize to put numbers to each seat. I use org chart and accountability chart interchangeably — accountability chart is more EOS language. But it's not just a name and three responsibilities. It's the number-one KPI that seat is responsible for. Whether human or agent, they're responsible for that number, and there needs to be tracking attached so you know simply red, yellow, or green — are we on track, off track, at risk? Having that data is so valuable.

If you had one piece of advice for agency owners right now, what would it be?

I'll tie it back to AI operations. The most important thing agencies can do is what we talked about: have an AI owner and an AI plan. Just get started on those. If the AI owner is you — the founder — lean in and embrace it, but be present that it's a responsibility you're taking on. You need to communicate it with the rest of the team. It's not something you do on weekends and then drop on the team Monday morning saying, 'I just flipped our entire service model with AI.' They need to be involved.

Make sure someone is in the AI ownership seat, then start crafting the AI plan. Get it out of your head and onto paper. We've been telling agency owners this since day one — whether it's AI, sales, or client success, get it out of your head and onto paper. That's the trap every agency owner falls into: they make grand assumptions that everyone knows what they know, but they haven't communicated it. Especially with how fast AI is moving, you have to verbalize it and share it. Get it out of your head, onto paper, build that plan, own that role — and you're 10 steps ahead of the next agency.

That's awesome. Thank you so much, Leah. This was a really good chat — so interesting and insightful. We'll have to do this again soon.

This is great. So much fun. I love chatting about these topics and I'm happy to answer any other questions people have on this space. I'm definitely not the only person in the AI ops space for agencies, but we're all learning as we go, so we love having these conversations and learning from what other agencies are doing too.