AI’s had a big week: Claude can design your pitch deck now, ChatGPT ads wobble, and more.

Figma might be about to become the fax machine of your tabs.

No designer? No problem: Claude Design lets anyone build prototypes, decks, and marketing assets just by describing what they need.

Anthropic’s new visual collaboration tool refines output through comments and direct edits, with your team’s brand system integrated from the start.

The use cases are genuinely useful:

• Landing pages, social assets, and campaign visuals
• Pitch decks exported directly to PPTX or Canva
• Interactive prototypes, no dev queue required

It also connects directly to Claude Code, shrinking the gap between design and build. If your team is perpetually waiting on a designer, this one’s worth a look.

Meanwhile, ChatGPT’s ad platform is barely holding together: Six-figure minimums, sky-high CPMs, and near-zero transparency. Cosy black box you’ve got there, OpenAI.

Sponsored retailers are already ranking higher in recommendations. Of course, this injects doubts that it may end up corrupting the answers. Hmm…

Consumer trust is the whole product here. If ads erode it, the platform eats itself. If they don’t, it becomes the most valuable ad surface in a generation.

For now, big brands with a big budget might find early testing worthwhile. Everyone else is better off watching how this plays out before committing.

And AI is everywhere. It just can’t read a clock: Stanford’s AI Index found generative AI hit 53% global adoption in three years, outpacing both PCs and the internet at the same stage.

Investment hit $581B in 2025, up 130% year-over-year. But the same models crushing PhD-level science questions still misread analog clocks half the time.

Google’s AIOs and AI Mode recommend different URLs for the same query 87% of the time. Two products, one company, not that much consensus.

Make of that what you will.

Search’s grip is slipping, but Google keeps tidying up its ad infrastructure

Did Jamie Lannister sneak in and kill the king?

The crown is losing its grip: According to IAB’s new report, search still commands $114.2B and nearly 39% of all digital ad revenue.

But its 11% growth looks modest next to social’s (33%), or video’s (25%) or programmatic’s (21%) climb.

The money’s not disappearing, though: Total US digital ad revenue hit a record $294.6B in 2025 with top 10 platforms controlling 84% of revenue.

If you’re not on the right channels, you could be paying more to fall behind.

Google noticed you hate manual tagging: The platform is testing a way to push conversion tags straight into Tag Manager. No more copy-pasting between setup and GTM.

Fewer steps, less room for error, faster implementation. If you’re managing multiple clients or complex tagging setups, this one’s worth watching.

Have your ID handy: Starting April 21, new Ads API authentications require a second factor. If your team regularly generates credentials, update your workflows before it hits.

Google Ads Editor, Scripts, and BigQuery Data Transfer all need attention. Existing OAuth tokens still work, so nothing breaks overnight, but new authentications follow the new rules from day one.

Service accounts are unaffected and worth switching to if you’re running automated workflows.

Five minutes now saves a broken workflow later.

These Google Ads mistakes tank your e-commerce campaigns

Expanding from paid social media to Google is less like opening a second location and more like opening a second restaurant…

…in a different country, with a different menu, where the customers behave nothing like the ones back home.

And Menachem Ani breaks down the six most common errors ecommerce brands make when crossing the channel divide.

1) You treat Google like a retention channel. Performance Max looks good early on. But months in, you may realize you’re mostly paying to capture purchases that were already happening.

New acquisition needs Shopping campaigns targeting unbranded keywords, not a retargeting tax on existing demand.

2) Not knowing how to get the most out of Google’s core levers. Social ads interrupt. Search ads intercept. Remember this distinction.

Upper and lower-funnel terms need separate campaigns. Your product feed needs constant attention. Sending high-intent cold traffic straight to a product page usually underperforms.

3) Letting operational issues interrupt campaign momentum: A billing hiccup that takes three days to fix doesn’t just cost you three days of ads. It costs you the recovery time after.

Broken pixels and feed errors are worse as they fail quietly, and so bad performance could get blamed on your strategy instead.

4) Building a campaign structure that’s too granular. Over-segmentation feels like control, but it isn’t. Smart Bidding needs conversion volume to work.

Spread thin across too many campaigns, it never gets enough data to optimize. Fewer, well-funded campaigns consistently beat fragmented ones.

5) Leaving campaigns on Max Conversion Value with no ROAS targets. Without a ROAS target, Google optimizes for volume, not profit.

Set a realistic target once conversion data exists, but don’t anchor it too aggressively too early, or you’ll starve campaigns before they learn.

6) Underfunding campaigns and keeping them stuck in learning. Google needs roughly 30–50 conversions before bidding stabilizes.

Underfund that window, and campaigns stay in learning indefinitely. You leave before the channel ever has a real chance.

Most of these mistakes share one root cause. Applying paid social logic to a channel that runs on completely different rules. You wouldn’t do that, would you?

Where Marketers actually use AI?

AI has become the Swiss Army knife of the modern marketing team.

But some “blades” are getting more use than others…

It appears that marketers are leaning on AI most heavily during the early stages of the content lifecycle.

Article topic research is the number one use case at 42%, followed closely by outline and brief development at 38%.

• Editing (36%): Using AI as a second pair of eyes.
• Drafting (35%): Getting the first 500 words down.
• Keyword Research (35%): Tied with drafting for SEO-focused tasks.

All these numbers show AI is firmly embedded in the “grunt work” of content production.

AI is excellent at “breadth” (researching and outlining) but still requires a human for “depth” (the actual final draft).

So it can’t be your creative director, but it is sure likely to be your creative assistant.

Use AI to generate 20 “out-of-the-box” topic ideas based on a single keyword. You’ll likely discard 18, but the remaining two could be angles you never would have considered.

AI can help your content workflow, but only if you know where to use it.

Pro subscribers get insights on what’s actually working, plus tactical playbooks for getting more out of AI without losing quality. $349/year for up to 10 teammates.