Prompt Horizon
Posts
Prompt Horizon - Issue #3: AI-Generated Creative

Prompt Horizon - Issue #3: AI-Generated Creative

Mastering AI in marketing, one prompt at a time.

Alex Lielacher
June 09, 2025

In partnership with

EDITOR'S NOTE

Hey there!

Welcome to the third issue of Prompt Horizon!

I'm Alex Lielacher, and this week I'm tackling a question that a lot of marketers think about these days: Should we trust AI-generated creative?

If you've played around with AI creative tools (admit it, you’ve all Ghibli-ed yourselves 😉), you've probably had that moment where you're genuinely impressed by what the machine spits out.

Maybe it was a headline that made you think, "Wow, that's actually pretty good!" Or perhaps it was an image that captured exactly what you were trying to convey.

But you've likely also had those moments where something feels... off.

The copy sounds generic. The image has that unmistakable AI vibe to it. Or worse, the facts are just plain wrong.

So, how do we navigate this new landscape? When should we let AI take the creative wheel, and when should we grab it back?

Let's find that sweet spot between machine efficiency and human intuition!

NEWS YOU CAN USE 📰

Gartner's latest report predicts that 30% of all commercial content will be AI-authored by the end of 2025, up from just 8% in 2023. However, they note that companies with human review processes report 42% higher audience engagement than those using fully automated workflows. [Source: Macrosoft Inc., May 15, 2025]

ElevenLabs has unveiled Eleven v3 (Alpha), which is described as their most powerful text-to-speech (TTS) model to date. The new model offers significant advancements over previous versions. This release represents a major milestone in AI-powered speech synthesis technology, providing creators and developers with unprecedented voice quality and control. [Source: ElevenLabs, June 3, 2025]

X platform bans third-party AI training with its content. Elon Musk's social media platform X has updated its developer policy to explicitly prohibit third parties from using its data to train AI models. According to The Verge, the new clause restricts developers from leveraging X's content or API to "fine-tune or train foundation/frontier models," tightening control over data usage. [Source: The Verge, June 5, 2025]

Microsoft launches Bing Video Creator. Microsoft has unveiled Bing Video Creator, a free tool leveraging OpenAI's Sora technology to generate short videos from text prompts. Unlike Sora, which is currently limited to ChatGPT Pro users, this release democratizes AI video production for all Bing users, mirroring the accessibility of Microsoft's earlier Image Creator tool. [Source: Winbuzzer, June 2, 2025]

THIS WEEK'S PROMPT 🤖

Actionable AI Prompt: Creative Quality Assessment Framework

Here's a prompt you can use with ChatGPT, Claude, or other AI assistants to develop a framework for evaluating AI-generated creative:

Act as a Creative Director with expertise in both traditional and AI-generated marketing content. Help me develop a comprehensive framework for evaluating when to trust AI-generated creative and when human intervention is needed.

Create an assessment rubric with the following components:

Brand Alignment (Score 1-10)

How well does the content reflect our brand voice, values, and positioning?

What specific elements align or misalign with our brand guidelines?

Audience Resonance (Score 1-10)

How likely is this content to connect with our target audience?

Are there cultural nuances or emotional elements the AI might have missed?

Creative Originality (Score 1-10)

Does this feel fresh and distinctive, or generic and formulaic?

Would this stand out in our competitive landscape?

Technical Accuracy (Score 1-10)

Are all facts, figures, and claims correct?

Are there any logical inconsistencies or flawed assumptions?

Strategic Alignment (Score 1-10)

How well does this support our current campaign objectives?

Does it effectively communicate our key messages?

For each component, provide:

Clear scoring criteria
Red flags that indicate immediate human review is needed
Examples of what excellent vs. problematic content looks like

Then, create a decision matrix that recommends:

Scores that indicate content can be used as-is
Scores that suggest minor human refinement
Scores that require significant human reworking
Scores that indicate the content should be rejected entirely

Finally, suggest a practical workflow for implementing this framework in a marketing team, including roles, responsibilities, and time estimates.

This prompt helps you create a structured approach to evaluating AI-generated creative, moving beyond gut reactions to a systematic assessment.

By scoring content across multiple dimensions, you can make more consistent decisions about when to trust the machine and when to trust your instincts.

The resulting framework will help your team establish clear boundaries for AI use, ensure quality control, and maintain brand integrity while still benefiting from AI's speed and idea generation capabilities.

Try running this prompt and then adapt the framework to your specific brand and marketing needs.

Farfetch Case Study: Using AI to Enhance Email Marketing 📈

The Challenge 🫷

Farfetch, a luxury online marketplace for fashion and beauty products, faced the challenge of improving the performance of its email marketing campaigns.

Specifically, they aimed to increase the open and click rates of their emails while staying true to the brand’s tone of voice. The primary goal was to use technology to optimize their email campaigns without compromising their brand identity.

The Solution 🚀

To address this challenge, Farfetch turned to Phrasee (now called Jacquard), a generative AI tool designed for enterprise-level applications. Using AI, the team enhanced various aspects of their email marketing strategy:

A/B testing and style optimization: Farfetch experimented with different phrases and writing styles to identify what resonated most with their target audience.
Optimized email subject lines: AI-generated subject lines were tailored for specific email categories such as promotional offers, abandoned cart messages, and wishlist reminders, improving the relevance and appeal of their content.
Personalizing email content: Farfetch adjusted the body of their emails to cater to the diverse preferences of their clientele and reflect the variety of brands that their marketplace offers.

Throughout this process, Farfetch ensured all AI-generated content aligned with their brand’s unique tone of voice, serving as a complement to their existing strategies.

The Results 📈

The integration of Phrasee (now Jacquard) into Farfetch’s email campaigns delivered impressive results:

Open rates

7% increase for promotional emails.
31% increase for event-triggered emails (e.g., abandoned cart reminders).

Click rates

25% increase overall.
38% increase for event-triggered emails.

These improvements highlight how effective optimization and personalization can enhance customer engagement and ultimately drive business goals.

Most importantly, in blind tests, audiences couldn't distinguish between campaigns created with the AI-human hybrid approach versus their previous all-human process.

Key Takeaways ✅

AI as an enhancer, not a disruptor: Farfetch demonstrates that AI shouldn’t replace a brand’s unique identity but instead work to enhance it. By maintaining their tone of voice and brand essence, they were able to boost email performance without compromising their core values.

Data-driven experimentation improves results: A/B testing phrases and styles helped Farfetch understand audience preferences, enabling data-backed choices to refine their email content.

Personalization at scale is key: With Phrasee (now Jacquard), Farfetch leveraged AI to craft personalized emails for a wide and diverse clientele, demonstrating how technology can cater to large, varied audiences effectively.

Impactful AI adoption delivers tangible results: The increases in open and click rates underscore the potential of AI when used strategically in marketing campaigns.

By embracing technology and remaining committed to its brand identity, Farfetch successfully enhanced its email marketing strategy, driving tangible improvements in customer engagement metrics. [Source: Data Feed Watch 2025]

Create How-to Videos in Seconds with AI

Stop wasting time on repetitive explanations. Guidde’s AI creates stunning video guides in seconds—11x faster.

Turn boring docs into visual masterpieces
Save hours with AI-powered automation
Share or embed your guide anywhere

How it works: Click capture on the browser extension, and Guidde auto-generates step-by-step video guides with visuals, voiceover, and a call to action.

Best part? It’s 100% free.

👉 Download the extension now

Wrapping Up 🌯

The most successful marketers aren't those who blindly trust AI-generated creative or stubbornly resist it—they're those who thoughtfully integrate it into their creative process while preserving human judgment where it matters most.

Think of AI as a creative partner rather than a replacement.

It excels at generating options, personalizing at scale, and handling repetitive tasks. But human creativity remains essential for strategic direction, emotional resonance, and that ineffable quality we call "taste."

The question isn't whether to trust the machine or your gut—it's knowing when to rely on each.

By establishing clear boundaries and systematic evaluation processes, you can harness the best of both worlds: the efficiency and scale of AI with the intuition and emotional intelligence of human creativity.

Until next week, I'd love to hear how you're balancing AI and human creativity in your marketing efforts.

Go forth and create—just don’t let the bots have all the fun!

Alex

PS: If you need help getting your brand to show up in AI Search, hit us up at Rise Up Media. We can sort you out!