MMarketing Against The Grain
← All frameworks
Marketing

AI Media Strategy Prompt Sequence

Turn one product name into an audience-led, single-medium campaign concept

Difficulty
Easy
Time to result
~days to results
Steps
8
Confidence
98%

The sequence converts a product name into a preliminary media strategy through ordered analysis and synthesis. First, the AI researches the typical buyer, the job they perform, and the problem the product solves. It then identifies content interests that both attract the buyer and maintain enough product relevance to support eventual monetization. Next, it examines which media the audience prefers and develops a list of suitable content topics. Rather than generating an unfocused omnichannel plan, the prompt chooses one medium and creates one differentiated campaign for it. Relevant creators can be incorporated when their audience and expertise strengthen distribution. The final output includes ideas for materially improving or scaling the campaign. Flanagan recommends iterating beyond the first response and treats this prompt as a minimum viable strategic prototype, not validated research.

Origin

Extracted from Marketing Against the Grain, where Kieran Flanagan builds and demonstrates a sequential media-strategy prompt, then tests it on Airbnb using Gemini Advanced.

Core principles

  • 01Begin with the product's monetizable target audience
  • 02Complete research steps in sequence before producing the campaign
  • 03Find content interests that overlap with purchase relevance
  • 04Choose one medium instead of spreading effort across all formats
  • 05Use AI output as an iterative starting point rather than a finished strategy
  • 06Match creators to both the topic and the intended audience

How to run it

  1. 1

    Define the Task and Input

    Tell the AI to create an engaging, differentiated media strategy for a product's target audience. Use the product name as the minimum initial input and require the steps to run in sequence.

    Pro tip Withhold the final output until the analysis steps are complete so the campaign reflects accumulated context.

    Watch out An underspecified task can produce a generic campaign unrelated to the audience or product.

  2. 2

    Analyze the Buyer

    Summarize the typical buyer, including their job and the problem the product solves. Store this as the working persona overview for later steps.

    Pro tip Separate distinct buyer groups when a product serves multiple meaningful personas.

    Watch out AI-generated personas are hypotheses and should not be represented as verified customer research.

  3. 3

    Find Monetizable Topic Overlap

    Identify subjects the buyer enjoys that also overlap with the product and could plausibly lead toward a sale. Explain why each overlap matters.

    Pro tip Prioritize topics where the product can contribute proprietary data, expertise, or a distinctive point of view.

    Watch out Broad interests with no commercial connection may attract an audience that will never become customers.

  4. 4

    Map Media Preferences

    Assess whether the buyer prefers blogs, newsletters, podcasts, YouTube, social video, or another medium. Use these consumption habits to constrain the campaign.

    Pro tip Validate the recommendation against existing audience analytics or customer interviews.

    Watch out Do not add formats simply because AI can generate them cheaply.

  5. 5

    Generate Content Topics

    Produce topic options based on the persona, monetizable overlaps, and media habits. Use them as campaign ingredients rather than isolated generic ideas.

    Pro tip Ask for increasingly differentiated revisions if the first topics resemble common industry content.

    Watch out Topics without a clear audience insight are unlikely to support a distinctive campaign.

  6. 6

    Choose One Medium and Campaign

    Select the single medium with the strongest audience fit and design one detailed media campaign for it. Specify the concept, content structure, audience appeal, and relationship to the product.

    Pro tip A focused campaign is especially useful for small teams that cannot sustain several channels well.

    Watch out The selected medium still requires human validation of cost, feasibility, and strategic fit.

  7. 7

    Integrate Creators Strategically

    Suggest creators with relevant audiences and explain a specific, distinctive role for each within the campaign. Use creator participation to strengthen both credibility and distribution.

    Pro tip Evaluate audience alignment and creative contribution, not follower count alone.

    Watch out Confirm brand fit, availability, rights, and commercial terms before treating a creator as part of the plan.

  8. 8

    Iterate and Improve

    Ask the AI to strengthen differentiation, detail, and engagement across several revisions. Investigate surprising suggestions rather than accepting or rejecting them automatically.

    Pro tip Compare multiple variants and preserve the strongest elements from each instead of blindly taking the final one.

    Watch out Repeated generation does not substitute for evidence from customers, creators, or campaign tests.

In the wild

Airbnb Unfiltered

Gemini analyzes Airbnb audiences, interests, and media preferences before choosing YouTube. It proposes partnering with travel vloggers to document unscripted Airbnb stays, including unexpected discoveries and mishaps. The scaling ideas include a contest for smaller travel creators and a hashtag inviting guests to share their own unfiltered moments.

The prompt produces a focused, creator-enabled campaign concept that Airbnb could investigate and refine.

Common mistakes

Stopping at the First Response

The initial AI output is a starting point; Flanagan recommends repeated iteration to make it more detailed, engaging, and differentiated.

Confusing Plausibility With Research

An idea such as Airbnb guests seeking home-design inspiration may be promising, but it remains an assumption until investigated.

Using Meaningless Improvement Targets

Asking for a vague “10x” improvement gives the model no concrete measure of what should become better.

Is it for you?

Best for

It is best for small marketing teams using AI to explore focused media campaigns for a known product and target market.

Not ideal for

It is not ideal for final investment decisions that require verified customer research, creator availability, budgets, and performance forecasts.

From the transcript

we always start with a task right we want to tell it what the task is

Kieran Flanagan · 30:30

the first thing we really want to do is research the products target audience

Kieran Flanagan · 32:30

choose one medium so I've choosen one because I want to actually do this for smaller companies companies who are trying to just like do…

Kieran Flanagan · 34:30

From the episode

I Created A Media Strategy For Airbnb With One Prompt (Multimodal Masterclass)