AI 168 – Weekly AI Marketing Newsletter

Edition: May 1, 2025

Hey Marketer,

Sometimes the most valuable AI applications aren't the flashy ones making headlines. This week, I found myself deep in the trenches, manually migrating a client's complex ad account after their previous agency left them in disarray. Three hours into what should have been a 30-minute task, I remembered the new Claude API connector I'd been meaning to try. One quick prompt later, my AI assistant parsed the entire messy export file, restructured it, and saved me at least 4 hours of tedious data work. The most impactful AI use cases are often the ones that save your sanity on an otherwise frustrating Wednesday afternoon.

🧠 This Week's Highlights

🗞 Top 10 News Items

  1. Salesforce Unveils Einstein 1 Platform with Major AI Upgrades – Salesforce released its complete Einstein 1 Platform upgrade, featuring Einstein Copilot, Studio, and Trust Layer for custom AI model building without code.
    Why it matters: Marketing teams can now train AI on their specific customer data without needing a data science team.
    👉 Read more
  2. Meta Quietly Launches Llama 3.1 – Meta released Llama 3.1, outperforming GPT-4 on certain benchmarks while remaining free for many commercial use cases.
    Why it matters: This is a game-changer for creating AI marketing applications without paying per-token fees.
    👉 Read more
  3. Canva Launches AI Brand Voice Generator – Canva has released a new AI Brand Voice tool that ensures consistent messaging across all marketing channels.
    Why it matters: This solves one of marketing's biggest challenges: maintaining consistent voice across teams and channels, with early users reporting 62% improvement in brand consistency scores.
    👉 Read more
  4. Google CEO Highlights AI's Role in Future Search Functions – During a federal court testimony, Google CEO Sundar Pichai emphasized the central role of AI, particularly Gemini AI, in the future of search functions.
    Why it matters: Marketers need to prepare for AI's increasing influence on search engine algorithms and advertising strategies.
    👉 Read more
  5. Duolingo Transitions to 'AI-First' Model – Duolingo announced plans to replace contract workers with AI to become an "AI-first" company, aiming to improve scalability in content creation.
    Why it matters: This shift underscores the growing trend of AI integration in content creation, impacting how marketers develop and manage content.
    👉 Read more
  6. OneOff Launches AI-Powered Fashion Search Platform – OneOff introduced a beta AI-powered fashion search platform that allows users to find outfits inspired by celebrities and creators.
    Why it matters: Highlights the potential of AI in personalizing shopping experiences and influencer marketing strategies.
    👉 Read more
  7. TikTok Invests $2 Billion in Creator Rewards Program – TikTok announced a $2 billion investment into its Creator Rewards Program for 2025-2026, aiming to boost creator-driven content on the platform.
    Why it matters: Significant investment in creators indicates a shift towards authentic, AI-enhanced content, affecting influencer marketing dynamics.
    👉 Read more
  8. MarketingProfs Highlights AI's Impact on UK Economy – MarketingProfs reported that Google estimates AI adoption could add £400 billion to the UK economy.
    Why it matters: Underlines the economic significance of AI, encouraging marketers to invest in AI skills and tools.
    👉 Read more
  9. Christopher S. Penn Shares Examples of AI Transformation – In his 'Almost Timely News', Christopher S. Penn shared five examples of AI transformation, illustrating the practical applications of AI in various industries.
    Why it matters: Provides marketers with real-world examples of AI implementation, offering insights into potential strategies.
    👉 Read more
  10. Diggity Marketing Reports on AI's Role in SEO Changes – Diggity Marketing's April 2025 roundup discusses how AI is influencing SEO practices, including Google's core updates and the evolving landscape of digital marketing.
    Why it matters: Understanding AI's impact on SEO is crucial for marketers aiming to maintain and improve their online visibility.
    👉 Read more

🛠 Tool of the Week: Keyhole

Keyhole is an AI-powered social media analytics tool that provides real-time insights and personalized recommendations for improving your strategy. It tracks engagement, reach, and sentiment while also offering competitor analysis to give your brand a competitive edge.
👉 Try Keyhole

💡 AI Tip of the Week: The "Graduated Specificity" Email Prompt

Looking to instantly improve your email marketing performance? I've been using this prompt structure with clients for the past few weeks:

Write an email about [topic] that progresses through these specificity levels:

  1. Universal problem (everyone in my audience faces this)
  2. Specific pain point (my exact target customer struggles with this)
  3. Example scenario (a concrete situation where this happens)
  4. Solution framework (the general approach to solving it)
  5. Specific action step (one thing they can do right now)

The email should be conversational, use short paragraphs, and include 2-3 questions for the reader. Subject line should create curiosity.

Brand voice: [describe your brand voice]
Target audience: [describe your audience]
Call to action: [what you want them to do]

This structure works because it starts broad (hooking a wide audience) then steadily narrows to speak directly to your ideal customer. I've seen this pattern increase open rates by 28% and click rates by 43% for a SaaS client.

The key is that it follows the natural decision-making process people use when deciding whether to engage with marketing content:

  1. Is this relevant to me at all? (Universal problem)
  2. Does this address my specific situation? (Specific pain point)
  3. Do they understand my exact circumstances? (Example scenario)
  4. Do they have a credible solution approach? (Solution framework)
  5. Can I take immediate action? (Specific step)

💬 From the DMs

Q: "Gabriel, I keep hearing contradicting advice about prompt length. Some say write detailed prompts, others say keep it simple. What's your take for marketing use cases?"
A: The confusion comes from conflating two separate issues: prompt complexity and prompt length. For marketing outputs, I've found that structure matters more than raw length. Focus on giving the AI these four critical elements:

This doesn't need to be 500 words. A focused 6-8 sentence prompt with these elements consistently outperforms rambling prompts with unnecessary detail.

🧠 Gabriel's Take

Meta's latest Llama 3.1 model is the sleeper hit of the year in AI for marketers. While everyone remains fixated on ChatGPT and Claude's latest features, Llama 3.1 is quietly becoming the backbone of custom marketing AI tools because of its permissive license and impressive performance. In six months, I predict we'll see an explosion of specialized marketing tools built on this foundation that solve very specific marketing problems at a fraction of the cost of the big API providers.

🎬 Fun Find of the Week

Look who's podcasting
Looking for a good laugh? A recent thread on X from @toolandtea features "AI babies hosting a serious podcast discussion" - and it's exactly as absurdly hilarious as it sounds. Something about tiny infants debating like seasoned pundits hits that perfect sweet spot of AI-generated weirdness.

AI babies hosting a podcast discussion

👉 Check it out on X

For marketers, this highlights how unexpected combinations can create highly shareable content. The virality of these AI-generated oddities demonstrates that humor remains one of the most effective ways to capture audience attention in overcrowded feeds.

📣 Know someone who'd love this?

Forward this or send them the link:
👉 https://trendfingers.com/ai168