21 March 2024

Will AI Automate 95% Of What Marketers Do?

2Frameworks

10Insights

Frameworks in this episode

Innovation6 steps

Assistant-Peer-Conductor Model

Match your operating model to AI's increasing level of autonomy.

Read →

Peak Performance6 steps

Compete Using AI

Automate repeatable work and reinvest your attention in distinctly human value.

Read →

Insights & moments

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

Myth Buster· 1

Myth Buster16:30

Today's AI Products Are in the Yahoo Era, Not the Google Era

Judging AI's long-term impact by current product limitations repeats the mistake of evaluating internet search through Yahoo's rudimentary web index. Flanagan argues that the category is still early and has not yet reached its PageRank-like breakthrough in product quality and experience.

The current wave was only about 16 months old in the episode
Early search engines looked unimpressive compared with Google
Current product weaknesses do not establish permanent limits
Agency and task disruption should be evaluated on capability trajectories

“it is a huge mistake to judge the impact of these tools based on the products that are out in the wild today”
Kieran Flanagan · 16:30

“this is the Yahoo era right”
Kieran Flanagan · 18:00

#ai adoption#innovation cycles#search history#technology

Hot Take· 3

Hot Take07:00

Vertical AI Could Turn Generalists Into High-Leverage Marketers

Specialized AI agents may invert the conventional advice that marketers must narrowly specialize. A strong generalist could organize agents trained for coding, design, analysis, and other vertical tasks while concentrating on goals, strategy, and integration.

Vertical AI targets specialized professional tasks
Agents can broaden the range of work one marketer can execute
Generalists may coordinate expertise they do not personally possess
Strategy and organization become important leverage points

“I believe vertical AI is where a lot of the disruption is going to come from in this next wave”
Kieran Flanagan · 07:00

“if you are just an incredible generalist you are able to work with these AI agents who are trained in vertical tasks you can do…”
Kieran Flanagan · 09:00

#vertical ai#generalists#agents#careers

Hot Take19:30

Bad AI Strategy Often Reflects a Weak Strategist

Flanagan challenges the belief that strategic thinking is automatically protected from AI. He argues that AI often reflects the user's own strategic knowledge, so generic prompts and weak marketing judgment predictably generate average strategies.

Strategy is not inherently immune to AI
User knowledge shapes the quality of generated strategy
Average strategists may blame the tool for average output
Domain expertise remains important for direction and evaluation

“the reason that most people think that AI cannot produce a great marketing strategy is because that person themselves cannot produce a great marketing strategy”
Kieran Flanagan · 20:00

“AI really reflects in a lot of CA places the knowledge you have about the thing that you were helping the AI to do”
Kieran Flanagan · 20:30

#strategy#prompting#marketing knowledge#ai quality

Hot Take31:30

AI Could Let Marketing Own the Entire Customer Journey

Flanagan predicts that AI-enabled marketers may eventually manage acquisition, sales warming, conversion support, customer marketing, and retention as one connected journey. Because marketing teams already have an automation mindset, they may be positioned to coordinate agents across these stages.

Marketing could expand beyond lead generation
Agents may automate work across acquisition and retention
A unified owner could improve customer-experience continuity
Automation may redefine the organizational boundary of marketing

“I believe in the future marketers could own entire customer Journeys”
Kieran Flanagan · 31:30

“we can actually make the entire customer experience incredible because we are automator at heart”
Kieran Flanagan · 32:00

#customer journey#marketing operations#automation#customer experience

Explainer· 3

Explainer06:00

What the 95% Marketing Automation Quote Actually Says

The disputed quote does not say that 95% of marketing itself will disappear. It says AI may perform 95% of what marketers currently use agencies, strategists, and creative professionals to do, implying that individual marketers could become substantially more self-sufficient.

The quote was attributed to Sam Altman but was not directly recorded
Its object is outsourced services rather than all marketing
The likely implication is greater marketer self-sufficiency
Fewer people may be needed for today's work even if new work emerges

“AI will do 95% of what marketers uh use agencies strategist and creative Professionals for today”
Kieran Flanagan · 06:30

“for me the key point in the quote that Sam was saying is what marketers use”
Kieran Flanagan · 07:00

#ai#marketing#automation#sam altman

Explainer11:00

Why Repeatable Agency Services Face the Most AI Disruption

Agencies and marketplaces are especially exposed when their services can be packaged as predictable tasks such as informational writing, translation, proofreading, editing, or clipping. AI offers clients better unit economics and can pass work between specialized agents, pressuring agencies built around average execution.

Packaged services are easier to automate
Informational content is more exposed than original creative work
Agent handoffs can cover drafting, proofreading, and revision
Expert-led AI-native agencies may replace account-management-heavy models

“agencies are 100% going to be completely disrupted”
Kieran Flanagan · 11:00

“AI is probably a better option for most marketers to Out Source their task 2”
Kieran Flanagan · 12:30

#agencies#services#automation#fiverr

Explainer22:30

Creative AI Can Produce C-Level Work but Still Needs Skilled Humans

AI tools already let resource-constrained marketers create credible design, video, and written assets. However, the episode places most unaided output around C-level quality and argues that skilled humans remain necessary to reach B- or A-level production standards.

Creative work was among AI's earliest disruption targets
Solo marketers can now produce more without specialist resources
Current output quality varies with the operator's content skill
High-end production still requires experienced creative professionals

“you should be using AI tools I think you can get like C- level content out C+ to C minus depending upon how you good…”
Kieran Flanagan · 23:00

“if you want to get to b or a level grade in then you still need a human to get you there”
Kieran Flanagan · 23:30

#creative ai#content production#design#quality

Story· 1

Story13:00

The Opus Clip Example Reveals AI's Short-Video Quality Gap

Jason Lemkin reportedly replaced an expensive, slow agency workflow with Opus Clip and received around 30 clips each week that he rated highly. Flanagan accepts the efficiency advantage but argues that quality must be judged against every creator competing for the same short-form attention, not merely against the displaced agency.

Opus Clip produced far more clips than the agency
The AI workflow offered stronger speed and unit economics
Short-form video is becoming exceptionally competitive
Internal quality grades may ignore the real attention marketplace

“he said he's is getting 30 a 30 or more Clips a week that he grades at like a minus from an app called Opus…”
Kieran Flanagan · 13:00

“I suspect they're like D+ C minus”
Kieran Flanagan · 16:00

#short video#opus clip#content quality#competition

Tool· 1

Tool23:30

Midjourney Character Consistency Solves a Practical Design Problem

Midjourney's character-consistency feature addresses a recurring production limitation: the same prompted character previously changed appearance across scenes. Preserving character identity makes image generation more useful for campaigns and narratives requiring visual continuity.

Earlier prompts struggled to preserve a character across settings
The feature allows a character to appear in multiple situations
Consistency is essential for practical design workflows
Rapid tool improvements are narrowing creative limitations

“it is called character consistency uh feature which basically means you can prompt mid Journey uh you can use that tag which means keep that…”
Kieran Flanagan · 23:30

“what you want when you are a designer is consistency of characters”
Kieran Flanagan · 24:00

#midjourney#image generation#design#character consistency

Takeaway· 1

Takeaway18:00

AI May Create Smaller Agencies Run by Exceptional Marketers

Rather than eliminating every agency, AI may enable boutique firms with a small number of deeply experienced marketers. These experts can train and operate agents that automate routine client work while they provide the judgment and expertise that average account-led agencies lack.

AI-native agencies can operate with smaller teams
Deep marketing expertise becomes more valuable
Agents handle repeatable customer work
The agency model may shift from labor scale to expert leverage

“we see these kind of AI specialist agencies run by a small group of incredible marketers with really deep expertise”
Kieran Flanagan · 19:00

“who know how to train the mo train the agents use the agents to actually automate a bunch of customers work”
Kieran Flanagan · 19:30

#ai agencies#business models#expertise#services