Why most marketing teams are using AI wrong — and what to use when.
The most common AI mistake we see from marketing teams isn’t using the wrong tool. It’s treating AI like one tool when it’s actually three distinct tiers, each built for a different kind of work, each replacing a different bottleneck.
We figured this out the hard way — by running it in parallel across a real engagement. Eight months embedded inside Bloomin’ Blinds, a window covering franchise with 60+ locations. By the end of it, median franchisee revenue had grown 41% year-over-year. The strategy drove that. The AI stack is what made the strategy executable at the speed and volume the engagement required.
Here’s what we learned.
Most teams only have Tier 1. That’s the problem.
When people say they “use AI for marketing,” they almost always mean they use it for writing. Content. Copy. Maybe some image generation. That’s real value — we’re not dismissing it. But it’s one tier of a three-tier stack, and treating it like the whole thing leaves two of the biggest productivity unlocks completely untouched.
The three tiers aren’t interchangeable. They don’t do the same job. And the order in which you build them matters more than most people realize.
Tier 01 · Content
Replaces: creative cycle time
Tools: Claude + Canva
This is the right place to start. Most accessible, fastest visible payoff, lowest barrier to entry for the team.
Claude compressed the brief-to-finished-asset cycle in ways that are hard to overstate until you’ve experienced it. On the Bloomin’ Blinds engagement, a strategic brief — audience, goal, tone constraints, word count, key messages — went in and a ship-ready franchisee email came out. Not a draft that needed heavy rewriting. Something reviewable and sendable. The cycle that used to take days ran in roughly 20 minutes.
The critical thing to understand about how this works: Claude didn’t decide what to say. The strategy, the audience insight, the tone judgment — all of that was human work done before the brief was written. What Claude replaced was the time between the decision and the finished asset. That’s the bottleneck it solves.
Canva handled design volume. For a 60+ location franchise network running simultaneously across five workstreams, the output requirement was significant — emails, social content, training materials, webinar visuals, a full franchisee marketing manual. The AI-assisted design tools in Canva let a small team produce at roughly 3x the throughput without sacrificing brand consistency. The creative director was still making every brand decision. AI removed the production bottleneck between the decision and the deliverable.
The limitation that matters: AI doesn’t make brand decisions. It doesn’t know your audience. It doesn’t have the strategic context that makes the difference between a communication that lands and one that gets ignored. The human judgment has to come first — AI just closes the gap between that judgment and finished work faster than any production process you’ve had before.
If you’re not using this tier yet, start here. Everything else builds on it.
Tier 02 · Operations
Replaces: human time on low-leverage work
Tools: Lindy + Claude Cowork + OpenAI Codex
This is the tier most teams skip. It’s also the one that gives senior people their time back — and that’s a fundamentally different kind of unlock than faster content.
Lindy ran the workflow automation layer on this engagement. Scheduling, routing, status checks, recurring follow-up — the repeatable tasks that don’t show up on a status report but quietly consume hours from people who should be doing strategy. Once those workflows are automated, that time doesn’t disappear. It goes back to the work only humans can do. That’s the compounding effect most teams miss when they think about AI ROI.
Claude Cowork and OpenAI Codex handled the file-level work. The clearest example: hundreds of marketing assets with inconsistent naming conventions needed to be sorted, renamed, and standardized across the entire batch. That’s a junior team member’s bad week — or a senior team member’s distracted afternoon, which is worse. With Cowork and Codex, it was done in hours. The team it freed up went back to strategy and client work.
This kind of work is invisible from the outside. But removing it from a human’s plate compounds across an engagement in ways that are hard to measure and easy to underestimate.
The limitation that matters: AI doesn’t fix a broken process. If your operations are disorganized, automation makes the disorganization faster. The workflow design has to come first. Map what you’re actually doing, identify what’s genuinely repeatable, then automate. In that order.
Tier 03 · Visibility
Replaces: developer cycles
Tools: Claude custom builds + HTML dashboards
This is the tier that tends to break people’s brains when they see it working — because most people don’t think of it as an AI use case at all.
On the Bloomin’ Blinds engagement, we needed real-time visibility for both the SIYL team and the client across every campaign component — workstream status, location-level activity, on-time delivery across 60+ franchisees simultaneously. The traditional path: write a ticket, get it into a sprint, wait five weeks.
Instead, we scoped it Monday and it was live Friday. Built in Claude. No developer. No sprint. Same week.
That’s not a party trick. It’s a structural change in what a small team can build and maintain. The dashboard gave the client a live view of what was moving and what wasn’t — not a monthly report, but actual real-time tracking. That changes how a client engages with the work. It also changes how a team manages a complex engagement when everyone can see the full board at once.
The limitation that matters: The build is easy. Knowing what to build is hard. Most teams want better visibility — fewer have thought clearly about what information would actually change how they make decisions. That’s the question to answer before you open Claude and start building. Custom builds also need maintenance. This tier is the last one to add, not the first, and it only earns its value when the content and operations layers underneath it are already working.
The Order of Operations
We ran all three tiers in parallel on this engagement. That’s what allowed a small, senior team to produce the output of a much larger marketing department. But if you’re building from scratch, the sequence matters.
Start with content. Fastest payoff, lowest barrier, clearest ROI. Every team should have this working before anything else.
Add operations once content is covered. This is where you reclaim senior capacity. The leverage tier — less visible than content, higher compounding value.
Add visibility last. It’s the most interesting one. It’s also the one that requires the most upstream clarity to be useful. A dashboard built on top of a broken content operation just gives you a faster view of the problem.
The instinct is to skip to Tier 3 because it’s the most impressive. Resist it. Build from the bottom.
Four Things Worth Writing Down
Match the tool to the task. Claude Chat for thinking and writing. Cowork for autonomous file-level work. Code-based tools for deployment. Three modes, three jobs. Using the wrong one wastes the specific advantage each one carries.
The spec is the product. The most valuable thing you produce in an AI-assisted workflow isn’t the copy — it’s the brief. The spec that makes every next step faster. Better inputs produce better outputs at every tier, and the discipline of writing a real brief before you open any AI tool is the highest-leverage habit you can build.
Build the artifact, not the description. Get something tangible in front of the team as fast as possible. Even rough. People align faster around something they can react to than something they have to imagine from a description.
The ceiling is a handoff signal. When a tool hits its limit on a task, that’s the cue to move up a tier — not to push harder against the constraint. Recognizing that signal is a skill. It gets faster with practice.
Ready to Build This for Your Team?
This is exactly what Train In Your Lane does — helping marketing teams and business operators build an AI stack that actually runs their operation, not just speeds up one part of it.
If you want the framework applied to your specific situation, we’d love to talk.
Leave a Reply