Last week, a founder in Singapore asked me to review her marketing plan. She had used an AI tool to generate it. The document was 12 pages long, full of phrases like "orchestrate a multi-channel brand narrative" and "leverage micro-influencers for authentic engagement." She had not executed a single item. She was stuck. This is the story of how we turned that textbook into a to-do list, and what happened when she followed it for 90 days.
The Plan Was Perfect. The Execution Was Zero.
Sarah (not her real name) runs a D2C brand selling a single hero product: a vitamin C serum for humid climates. Her target is women in Southeast Asia, aged 28-45, who have tried Korean skincare but found it didn't hold up in the heat. She had 80 customers from her launch six months prior. Growth had flatlined.
Her AI-generated plan was, on paper, flawless. It identified her target audience, suggested channels (Instagram, TikTok, SEO blog), and even proposed a content calendar. The problem was in the verbs. Every instruction was strategic, not tactical. "Develop a content strategy that highlights product efficacy." What does that mean on a Tuesday at 2 PM?
As the original review pointed out, this is the universal flaw. The AI gave her a university assignment, not a Monday morning task list. Her plan had no connection to her actual resources: her own time, a $500 monthly ad budget, and a phone for taking photos.
We Threw Out the Strategy Doc and Built a Triage List
We started with a brutal triage. We opened her 12-page plan and highlighted every sentence that contained an abstract verb. Leverage. Develop. Implement. Optimize. We replaced each one with a question: "What is the very next physical action?"
The plan said "Leverage Instagram Reels for demo content." Our question: What is the first Reel? She said she could film her applying the serum on her own skin, talking about the sticky feeling of other serums in Singapore's humidity. Good. Action: Film that one Reel. Not "create Reel content." Film *this* Reel.
The plan said "Optimize Google Business Profile." Action: Go to GBP right now. Upload these three photos: the product bottle, the ingredient list, a screenshot of a five-star review. Write one post about the serum's formulation for heat.
In 45 minutes, we had a list of 14 specific, atomic actions. None took more than 30 minutes to complete. This was her first two weeks. This was the plan.
The 90-Day Framework: One Channel, One Metric, One Action
We agreed on a simple rule for the quarter: one primary channel, one core metric, one daily marketing action. She chose Instagram as her primary channel. Not Instagram, TikTok, LinkedIn, and a blog. Just Instagram. Her core metric was not followers. It was profile visits from non-followers—a proxy for curiosity driven by Reels.
Her one daily action was to post one Story featuring a customer testimonial or a raw, unscripted thought about skincare in humidity. Every single day. This took her less than five minutes. It was not about production quality. It was about consistency and building a recognizable, reliable presence.
The weekly cadence looked like this:
- Monday: Film and post one Reel (product demo, myth busting, routine walkthrough).
- Tuesday: Engage with 10 relevant comments on other skincare creators' posts.
- Wednesday: Check Google Business Profile insights, respond to any new reviews.
- Thursday: Send a personal thank-you DM to any new customer who followed from an order.
- Friday: Review the week's profile visit metric. Plan next week's Reel topic.
This was not in the AI plan. The AI plan had a "content calendar" with themes like "Wellness Wednesday" and "Transformation Thursday." It was generic. Ours was specific to her product, her bottleneck (time), and her goal (establishing trust).
What the Numbers Said After 90 Days
She executed this simple, tedious, unglamorous plan for 13 weeks. Here is what changed.
Profile visits from non-followers increased by 300%. Her follower count grew by 1,200, but that was a vanity metric. The important shift was in DMs. She went from receiving 2-3 product questions per week to 10-15. These were warm leads, already engaged.
Website traffic from Instagram tripled. Sales attributed to Instagram (using simple UTM codes) increased by 40%. Her email list grew from 200 to 550 names, purely from a link in her profile visited by people who saw her consistent Stories and Reels.
The most telling result was what didn't happen. She did not run a single paid ad. She did not collaborate with macro-influencers. She did not produce a single piece of "hero content." She executed a low-friction, high-consistency tactical plan built from the rubble of a strategic AI document.
Why This Works When AI Plans Fail
The AI was not wrong. It was just several layers of abstraction too high. It is excellent at outlining the *what*. It is terrible at defining the *how* for a specific human with specific constraints.
Sarah's success came from inverting the AI's process. The AI starts with a broad goal ("increase brand awareness") and works down to generic tactics. She started with her specific capacity (30 minutes a day) and worked up to the smallest possible actions that could accumulate into a result.
This is the gap no AI marketing plan generator has bridged. They cannot model your personal context—your aversion to being on camera, your skill with Canva, your network of three friends who will share your first post. They give you the map of the territory. You still have to walk the path, one step at a time. The winning approach is to use the AI for the map, then immediately break the map down into footsteps.
What we'd do next
If you are staring at a beautiful, useless AI marketing plan, do this one thing today. Open it. Find one strategic recommendation. Ask: "What is the first physical action this implies?" Do that action. Then do the next one. The plan that gets executed is better than the perfect plan that stays in a PDF. For founders who want this principle applied to sales outreach—turning strategy into sendable emails—you can give MiraReach a try.
— Mira