Power-packed AI personas 🤸

Hello marketers. Welcome to AI Marketing School, where we dish out the latest and greatest in AI-powered marketing.

In this week’s issue:

1. AI Marketing Update: AI marketing: a year in review.

2. The Stack: Functional, ethical AI personas.

Onwards!

AI MARKETING UPDATE

Another blockbuster year for AI

It’s been a massive year for AI, though perhaps the technology is now following a more traditional path of slower technological iteration rather than being as explosive as some predicted. 

Nevertheless, we’ve seen plenty of milestones smashed and we’re set for another exciting year in 2025.

From a marketing perspective, there’s been some definitive changes with AI-driven search moving from concept to reality and AI-generated imagery creating big controversies in the advertising world. 

Here’s 2024 in review:

1. Search gets a complete makeover

2024 brought about massive changes to search. 

First came Google's massive reindexing effort in March, wiping countless AI-generated pages from search results overnight. Many have since recovered, but spammy AI writing remains a risky proposition for both existing and new sites. 

Then the floodgates opened for generative AI search. 

ChatGPT Search arrived in October, accompanied by Google's expanded SGE rollout. Billions of internet users worldwide are receiving complete answers to their search queries, synthesized from multiple sources, with direct quotes and embedded media.

This triggered a complete rethink of content strategy. Research from earlier in the year showed shorter articles around 800 words were actually performing better. 

Google started surfacing what they called "hidden gems" – content with first-hand experience and authentic insights.

Smart marketers are creating clusters of shorter, more specific content that provide real value at each stage of the user journey. Consider offering genuine expertise and original perspectives that AI can’t replicate.

As Gartner predicts, brands could see a 50% drop in organic search traffic by 2028 as users shift to AI search tools. 

Our collective response? Focus on "knowledge gain" – creating content that adds something new to the conversation rather than just rehashing what's already out there. 

Original research, expert insights, and authentic experiences are all the more important going forward.

2. A wild year for AI ads

We saw fascinating experiments with AI advertising this year. Coca-Cola closed out 2024 with their AI Christmas ad that became infamous for truck wheels that didn't actually spin. 

Google had to pull their Olympics campaign after backlash over AI-written fan letters.

But it wasn't all misses. Tech company Atera saved a million dollars with AI-generated ads that embraced creative absurdity. 

And the Met Gala showed AI's viral potential when generated images of Katy Perry, Lady Gaga and Rihanna fooled millions – including Perry's own mom.

Some celebrities even started embracing the technology – like Singaporean star Jamie Yeo signing deals to let companies use her AI-generated likeness in marketing campaigns.

Expect more AI-human collaborations as we descend deeper into the uncanny value through 2025. 

3. Politics gets complicated

Politics and AI made for an interesting mix this year.

Trump's team sparked controversy with deep fake images of Taylor Swift supporters, while an AI-generated Biden robocall in New Hampshire was foreboding for how AI will interact with politics going forward.

There are positives, too. The AAPI Victory Alliance used AI to translate campaign messages into six Asian languages in record time, making political content accessible to millions more voters.

4. Tools break new ground

2024 was the year AI tools reached new heights.

• OpenAI's Advanced Voice Mode enabled natural conversations with emotional depth

• ElevenLabs launched Audio Native, revolutionizing voice content

• Developers created innovative tools for AI storytelling and video production

• Voice cloning opened new marketing possibilities while raising ethical questions

• Just out – Google’s Gemini 2.0 is offering unprecedented multi-modal functionality

• Also just out - OpenAI announces a ‘video mode’ where the AI can interact with real-time images via your camera.

The models kept advancing.

• Claude 3.5 Sonnet smashed benchmarks

• OpenAI’s GPT-4o arrived with faster processing

• GPT-o1 introduced more sophisticated analysis, a hint of how more agentic forms of AI will become more common

Workflows became more efficient.

• ChatGPT added memory features 

• Broader AI integration with image generators simplified creative processes

•  Anthropic’s Projects and Artifacts added great functionality to their tools

5. The birth of LLMO

Want to prepare your content and marketing strategies for the AI era? 

Keep an eye out for “LLMO” – Large Language Model Optimization. 

A Harvard study demonstrated how strategic content can tangibly influence AI recommendations.

The coffee maker study revealed how carefully crafted product descriptions could alter AI rankings and recommendations. This reveals new strategies for content optimization:

• Strategic Text Sequences to improve AI visibility

• Optimized product descriptions for AI comprehension

• New techniques for content relevance in AI systems

• Integration of traditional SEO with AI optimization

Keep your ears to the ground to learn more about LLMO and how you can gain an edge as marketing shifts to leverage AI as a marketing channel. 

What's next?

Predicting AI’s trajectory is somewhat of a fool’s errand, but AI models will definitely become more intelligent, multimodal, and autonomous – capable of creating and executing plans.

Native AI functionality in mobile devices combined with AI search will change how information is created and shared. 

The traditional internet really does feel vulnerable now.

Marketers will need to experiment and challenge old tactics if they’re to succeed going forward. 

THE STACK

Leveraging AI personas

Marketing personas – the carefully crafted personalities brands adopt and target – have come a long way from the basic customer profiles of the past.

Back then, personas were pretty simple: create a fictional "ideal customer" with basic demographic details and use that to guide your messaging. 

Maybe Sarah, 35, professional mom who likes yoga. Or Mike, 28, tech-savvy gamer who orders lots of pizza.

Today, AI has transformed how we create and manage marketing personas. 

You can use AI to totally transform your communications and marketing messaging, convincingly adopting an identity you have little knowledge or understanding of — at least superficially.

Some marketers have been taking this in some ethically dubious directions.

Without naming names:

• LinkedIn sales teams are reporting significantly higher response rates using AI-generated female profiles for cold outreach

• A business owner on Reddit found customer service dramatically improved when using a feminine name - with customers being more polite, more likely to leave reviews, and even apologizing for asking for discounts

• Marketers are seeing faster social media growth with AI personas, particularly in tech-focused areas like cryptocurrency

• One marketer increasing conversion rates from 1.9% to over 7% using an AI-generated photo of an attractive female.

Beyond gender, marketers are using AI to impersonate different types of voices — from corporate professionals to Gen Z influencers.

That’s nothing new in itself, but it’s more that the realism is much greater than it once was, which makes it all the more deceptive.

As a ‘growth hack,’ people might view it effective to use AI to impersonate who you want to get what you want. But there are some massive risks. 

The risky business of persona deception

Pretending to be someone you're not is ethically questionable at best. 

Some will say “all is fair in love and war,” and argue that brands have used dubious psychological tactics to get into our heads (and wallets) for centuries.

But AI impersonation takes this to another level. Unlike traditional tactics, it doesn’t nudge or suggest — it pretends, borrowing someone else’s identity to convince you.

It’s one thing to persuade. It’s another to lie convincingly, en masse, at the push of a button.

Beyond moral implications though, there are practical risks:

• Getting caught can devastate brand reputation. Just imagine the headlines: "Company caught using fake female personas to manipulate customers.”

• In an era of increasing privacy concerns and online scams, people are getting better at spotting fakes. Being outed as inauthentic can permanently damage customer trust.

• Even if it works short-term, deceptive practices often backfire when scaling. What starts as "harmless" A/B testing can evolve into systematic manipulation.

• There's growing backlash against gender-based marketing manipulation, especially from younger audiences who value authenticity and transparency.

• The legal situation around AI-generated personas and digital identity is still evolving — what's technically possible might not stay legally permissible. Manipulating customers with AI is illegal in many jurisdictions.

Building better AI personas

Here's what actually works for creating effective AI personas without crossing ethical lines:

Start with mixed methods analysis

Start with the humble Google Analytics — it's still the backbone of understanding user behavior.

Look for patterns in how different segments interact with your content, which topics drive the most engagement, and how behavior varies between new and returning visitors.

Pay special attention to content performance across different audience segments.

Next, dive into native social analytics across platforms. Each network offers unique insights:

• LinkedIn's content analytics reveal what resonates with professional audiences

• Meta's detailed engagement metrics show how tone affects response rates

• X’s conversation data highlights natural language patterns and trends

• Instagram's story metrics demonstrate visual preferences

• Use Meta AI to search for trending content

From there, you can feed your data into AI to understand your audience and develop effective communication personas (not fake identities) to boost your marketing efforts.

Define clear voice parameters and create AI systems

With your data insights in hand, it's time to build AI-enhanced marketing communication and content systems.

Collect information from the business

Gather details about the business, its audience, and communication goals. Use the following structure to organize this data:

• Who we are: Describe the company, its mission, and its core offerings. For example, “We are EcoSphere, a sustainable home technology company creating products like smart thermostats and solar-powered air systems to make eco-friendly living accessible.”

• Who we serve: Define the customer segments, such as “Sustainability advocates who are passionate about reducing carbon footprints” or “Tech-savvy early adopters excited by smart home innovations.”

• What we’re trying to achieve: Outline objectives, such as “Build thought leadership on LinkedIn, increase Instagram engagement, and drive conversions via email campaigns.”

• Products: Any relevant product info.

Define tone and style for each platform

Specify how the AI should communicate for each channel. For example:

• LinkedIn: Professional and insightful.

• Instagram: Casual, engaging, and visually driven.

• Email campaigns: Persuasive and action-oriented.

Write sample prompts and outputs

Provide examples of prompts and expected responses to guide the AI. For example:

• Prompt: “Write a professional LinkedIn post about EcoSphere’s solar-powered air system.”

• Output: “EcoSphere’s Solar Air System reduces household energy use by 30%, helping homeowners save money and lower their carbon footprint. Learn more: [link].”

Combine all content into a single document

Format everything into a single, clear document that the AI can process.

Here’s an example built with a Claude Project:

The goal here isn't to create perfect fake personas — it's to be a more consistent, responsive version of your authentic brand. 

Find that sweet spot between AI efficiency and human authenticity, using AI to enhance your knowledge and intuition, not fabricate fake personalities to actively manipulate people.

Hope you enjoyed this week’s issue. If you missed it last week, you can read it here.

If you found it useful, please recommend it to a friend or colleague.

Until next time. Happy marketing.

—The AI Marketer