07 May 2024

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

2Frameworks

9Insights

Frameworks in this episode

Marketing8 steps

AI Media Strategy Prompt Sequence

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

Read →

Marketing6 steps

Multimodal Content Readiness Framework

Build platform-native content skills before search disruption forces the shift

Read →

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Myth Buster· 1

Myth Buster22:00

A Hit Podcast Episode Is Not Automatically a Hit YouTube Video

Flanagan compared the show's YouTube and RSS performance using OpenAI's Code Interpreter. Almost no episodes performed strongly on both platforms, illustrating that visual and audio audiences reward different subjects and packaging.

Cross-platform performance overlap was nearly nonexistent
YouTube is a visual discovery and viewing experience
RSS podcast consumption is primarily audio-driven
Identical content should not be expected to perform equally everywhere

“what episodes performed great across each platform you know the sad thing none there was literally one they are so different”
Kieran Flanagan · 22:30

“what performs well on YouTube does not perform as well an RSS”
Kieran Flanagan · 22:30

#youtube#podcasting#analytics#content performance

Hot Take· 2

Hot Take11:00

Stable Google Rankings Can Conceal Falling Search Demand

Flanagan says companies with strong organic strategies are seeing informational keywords retain their rankings while delivering less traffic. He interprets this as early evidence that AI is absorbing demand for generic educational content even before conventional ranking reports show a crisis.

Rank stability does not guarantee stable traffic
Informational searches face particular exposure to AI answers
Generic educational content is easy for AI to reproduce
Marketers should monitor actual volume and conversions, not rankings alone

“the keyword rank can remain consistent but actually get in much less volume over time”
Kieran Flanagan · 11:30

“educational content the content we have created traditionally to acquire users and customers is going to be heavily disrupted”
Kieran Flanagan · 11:00

#seo#organic traffic#informational keywords#search disruption

Hot Take12:00

YouTube Is This Generation's Television

Flanagan and his co-host view YouTube as a likely beneficiary if traditional web search loses attention. Brands may need YouTube channels, but video is harder to produce, grow, and monetize than search-led text content.

YouTube may gain attention as conventional search changes
Brands increasingly need durable video distribution
Video competes directly against entertainment for attention
Production and monetization are harder than in search

“YouTube Kip and I believe that YouTube is this generation's TV right it will disrupt TV”
Kieran Flanagan · 12:30

“YouTube is much harder to grow video is much harder to do and it's much harder to monetize that demand”
Kieran Flanagan · 12:30

#youtube#video marketing#brand media#attention

Explainer· 3

Explainer00:30

The Four Pillars of the AI-Era Marketing Playbook

Flanagan argues that marketers must develop the capabilities companies will still value after AI automates routine work. He identifies four pillars: multimodal content, personalization at scale, creator-led strategies, and the collapse of more of the customer journey into marketing.

Routine marketing work may become heavily automated
The most valuable marketers will own scarce strategic capabilities
Multimodal content is one of four proposed pillars
Personalization, creators, and customer-experience ownership complete the model

“there's four things four pillars that will really matter in that era”
Kieran Flanagan · 02:00

“there's something called a multimodal Content strategy there is personalization at scale there is Creator Creator L strategies and then there's the collapsing of the…”
Kieran Flanagan · 02:00

#ai marketing#automation#marketing careers#customer journey

Explainer03:30

The Bull and Bear Cases for AI Disrupting Google Search

Flanagan examines conflicting evidence about whether AI answer engines are disrupting conventional search. Google's North American market share appeared to dip while Bing rose, but query volume and Google's daily user retention remained strong compared with OpenAI's drop-off.

StatCounter data showed a notable Google market-share dip in April 2024
Bing may have benefited from moving earlier on Copilot
Search query volume remained comparatively stable
Google retained daily usage better than OpenAI chat products
The evidence supports monitoring rather than assuming one inevitable outcome

“people are starting to really prefer an answer Engine versus a search engine”
Kieran Flanagan · 06:00

“these AI chat Bots really need to figure out how to become much stickier for these users”
Kieran Flanagan · 08:30

#google#search#answer engines#market share#ai

Explainer23:30

Why Marketers Must Learn to Defend Indirect Channels

Search offered unusually clear attribution from content through traffic, customers, and revenue. Video, social, newsletters, and podcasts often have less direct attribution, so marketers must combine measurement with a persuasive explanation of why audience-building work deserves investment.

Search made individual marketing contributions unusually visible
Growing media channels often influence rather than directly capture demand
Privacy and tracking changes further weaken direct attribution
Views and engagement can be challenged by founders and executives
Strategic communication becomes part of the marketer's job

“we are going into an era where there's going to be much more indirect”
Kieran Flanagan · 26:00

“your job is to have a Clear Vision be able to articulate that Vision why it matters”
Kieran Flanagan · 26:30

#attribution#marketing measurement#executive communication#privacy

Story· 1

Story36:00

Gemini Builds an Airbnb YouTube Campaign From One Prompt

The live demonstration asks Gemini Advanced to analyze Airbnb's audience, interests, preferred media, and potential topics. It selects YouTube and proposes “Airbnb Unfiltered,” a creator campaign centered on authentic, unscripted stays and unexpected moments.

Gemini identifies travelers, budget explorers, and digital nomads
Suggested interests include destinations, sustainable travel, and home design
The model selects YouTube for its visual fit with travel
The proposed campaign uses established travel vloggers
Contests and user-generated content could extend distribution

“the campaign it has here is Airbnb unfiltered”
Kieran Flanagan · 39:00

“this campaign partners with popular travel vloggers to create raw authentic videos documented their Airbnb stay”
Kieran Flanagan · 39:30

#airbnb#gemini#campaign ideation#creator marketing#youtube

Tool· 1

Tool13:00

AI Tools Are Removing the Production Bottleneck

AI-assisted voice, video, and editing tools make formerly resource-intensive formats accessible to smaller teams. Flanagan highlights text-based podcast editing and synthetic video workflows as major reductions in production effort, even though current avatars can still feel monotone.

Audio and video historically required specialist resources
Text-based editing can dramatically accelerate podcast production
Generative video and avatars reduce recording overhead
Current synthetic presenters still lack realism and emotion
Tool improvements can unlock more marketer creativity

“the first time I used the script to edit a podcast like it's a it's a 10x Improvement”
Kieran Flanagan · 14:00

“AI will actually unlock a lot of marketers creativity”
Kieran Flanagan · 13:30

#ai tools#video production#audio production#content creation

Takeaway· 1

Takeaway29:00

A Prompt Can Be the Minimum Viable Version of an AI Product

Flanagan uses prompting as a low-cost way to test an AI workflow before building an interface or application. Writing the prompt also forces the marketer to articulate the task, inputs, sequence, and desired decisions clearly enough for another intelligence to execute them.

Prompts can test product ideas before software development
Prompt writing exposes vague strategic assumptions
A successful workflow can later become an app or agent system
Complex sequences may ultimately work better as specialized cooperating agents

“it's a great minimal viable way to like test something in AI”
Kieran Flanagan · 29:00

“the prompt teaches you how to like just build that articulate the thing you were trying to do”
Kieran Flanagan · 30:00

#prompt engineering#prototyping#ai products#agents