13 May 2025

This AI Replaces Your Marketing Team in 30 Minutes (Step-by-Step)

2Frameworks

8Insights

Frameworks in this episode

Innovation5 steps

Five-Step AI Capability Ladder

Progress from basic prompts to custom AI-powered applications

Read →

Marketing4 steps

Three Short-Form Content Archetypes

Classify each idea by the response it should create

Read →

Insights & moments

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

Hot Take· 2

Hot Take21:30

AI May Unlock Creativity Instead of Killing It

Kieran rejects the assumption that generative AI necessarily diminishes creativity. He sees a well-trained assistant as a way for individuals to execute more of their ideas and explore possibilities that were previously outside their practical reach.

Creative assistance can expand what one person can produce
AI can help execute ideas beyond a user's existing technical abilities
Teaching the assistant preserves the human's role in directing the work

“it is the best onl for creativity which is kind of intuitive spicy take to what most people are think of AI or worry about…”
Kieran · 21:30

“I actually think it unlocks the creativity within us because you now can do much more”
Kieran · 21:30

#creativity#generative ai#augmentation

Hot Take26:30

Buying Decisions Are Moving toward Word of Machine

Watching an AI coding tool choose languages, hosting, and deployment led to a broader prediction about machine-directed adoption. As assistants increasingly recommend technologies and services, products may need to earn selection by AI systems as well as recognition from people.

AI systems increasingly select components on a user's behalf
Machine recommendations may shape product adoption
Marketing may need to account for AI-mediated buying decisions
Technical defaults can become powerful distribution channels

“we're moving from a word of mouth to a word of machine world”
26:30

“AI is going to direct a lot of the adoption and buying decisions in the future”
26:30

#ai recommendations#distribution#buying behavior#product adoption

Explainer· 1

Explainer25:00

When an AI Agent Should Become an App

The hosts distinguish narrow process automation from workflows that require a full application. Agents work well for specific repeatable processes, while apps become useful when several data sources and more complex interactions must be coordinated.

Use agents for narrow and repeatable process automation
Move toward an app when multiple data sources are required
Complex workflows may need interfaces, storage, and deployment
Do not assume every useful AI workflow needs full software

“agents are really good at automating like very specific processes”
25:00

“you kind of graduate to needing an app when you need lots of different data sources what you're trying to do gets very complex”
25:00

#ai agents#applications#automation#software

Story· 1

Story25:30

The App That Ranks the S&P 500 by Marketing Strength

One host describes building an application that combines the S&P 500 with publicly available marketing data, primarily from Similarweb. The proposed system would score companies on marketing strength and examine whether stronger marketing correlates with superior business growth.

Combine public company lists with third-party marketing data
Design an algorithm that scores marketing performance
Track how highly ranked companies grow over time
Use AI coding to pursue projects previously blocked by technical skills

“I want to build an algorithm of that marketing data and rank the S&P 500 based on how good they are at marketing”
25:30

“what's crazy is that I could have never done that before”
26:00

#s&p 500#marketing data#similarweb#ai coding

Tool· 1

Tool22:00

Turn Forgotten Content Archives into Learning Data

SlideShare and YouTube are presented as underused stores of expert knowledge. Existing presentations, videos, and transcripts can supply material for studying successful methods and generating modern content derivatives.

Look for archives containing concentrated expert knowledge
Use transcripts and PDFs as machine-readable source material
Add new analysis or modern context rather than merely reposting old work
SlideShare remains a potentially valuable but neglected source

“where is there a treasure Trove of data and learning that people may not be using or getting leverage from”
Kieran · 22:00

“SlideShare is a great example”
Kieran · 22:00

#slideshare#youtube#content archives#learning

Takeaway· 3

Takeaway08:30

Deep Domain Expertise Becomes More Valuable with AI

Kieran argues that strong AI results come from teaching the system the craft knowledge accumulated by an expert. Rather than making expertise irrelevant, AI can amplify people who understand a domain well enough to explain and encode it.

AI output reflects the quality of the expertise supplied to it
Experts can teach assistants to replicate recurring parts of their craft
Weeks of domain-specific training may be required for strong results

“your results with AI a reflection of you as a manager of how you were managing that AI assistant”
Kieran · 08:30

“it's going to be a boom for people who really understand the craft have deep domain expertise”
Kieran · 09:00

#domain expertise#ai training#craft#knowledge

Takeaway16:00

Stop Debating Customer Tone and Test It

The hosts challenge subjective claims about which tone customers prefer. AI-generated variations make it practical to run structured tests across styles that were previously difficult for small marketing or sales teams to produce at scale.

Treat beliefs about preferred tone as testable hypotheses
Generate controlled variants across several writing styles
Use performance data to determine which style works
Separate subjective internal opinions from customer behavior

“you can say hey we all we have these four different types of tone voice everything that we think our customers like and we're just…”
16:00

“you can actually run very structured tests on unstructured information”
16:30

#experimentation#tone of voice#email marketing#sales

Takeaway27:00

AI Turns Impossible Problems into Really Hard Ones

The episode closes with a mental shift attributed to Microsoft's CTO: AI repeatedly moves problems from impossible to difficult. The strategic opportunity is to identify consequential problems that have just crossed that boundary and attempt the hard work competitors will avoid.

Model improvements continually expand the set of feasible problems
Hard problems can create differentiation because most people avoid them
Select an impactful problem rather than building novelty for its own sake
Reassess previously impossible projects after major model advances

“the best way to think about AI is it makes the impossible problems just really hard”
Kieran · 27:00

“most people will not do the hard thing”
Kieran · 27:30

#innovation#problem selection#competitive advantage#ai models