AI Field Notes – Issue 13

What we’re seeing, thinking, and building in the AI x Marketing trenches.

1. Speed is strategy now.

The agencies that will win in AI aren’t necessarily the smartest or the biggest—they’re the fastest.

No more waiting 6–12 weeks to build a productized marketing tool. If you can’t build and deploy a working solution—like a calculator, quote engine, or diagnostic tool—in 48–72 hours, you’re behind.

Speed isn’t just a competitive advantage anymore. It’s the barrier to entry.

→ Velocity is the new value.

2. Productized solutions > traditional services

Clients don’t just want websites or ad campaigns. They want solutions to real problems.

We’re leaning into building “solution-products” like:

• Instant quote generators
• Product spec finders
• Troubleshooting workflows
• Dynamic pricing tools
• GPT-powered support agents

These aren’t gimmicks. They’re digital products that live on a client’s site and solve customer problems directly.

→ Marketing teams need to act more like product teams.

3. Search isn’t evolving—it’s eroding

We’re watching the slow erosion of traditional search behavior.

Yes, people still “Google,” but the numbers don’t lie. Attention is moving toward problem-first interactions via chatbots, GPTs, embedded tools, and social platforms.

Saying “search is fine” is starting to sound like telling people in 2005 that newspapers are coming back.

→ Marketers who deny this shift are doing their clients a disservice.

4. Custom GPTs are becoming the new business OS

Most SaaS tools are just GPT wrappers. Useful, yes—but generic.

The future? Companies building their own GPTs, trained on their workflows, data, tone, and logic.

Custom GPTs won’t replace everything. But they will become the connective tissue of an AI-enabled org—acting as copilots, advisors, and execution engines.

→ You don’t fit your business to a tool. You fit the tool to your business.

5. Vibe coding is real—and powerful

Vibe coding (yes, that’s what it’s called) is messy but promising.

It’s the art of building workflows through natural language, not syntax.

It’s still early and full of bugs—but for those who understand process design, it’s a massive unlock. The coders of the future aren’t engineers. They’re operators who understand systems and workflows.

→ Know the flow, then prompt the code.

6. The gap is closing fast

The market is being flooded with “AI solutions”—but most of them are still surface-level.

Here’s the problem:
Companies only have a short window to get this right.

Soon, everyone will have access to the same tools. The edge will belong to those who customize and integrate before the window closes.

→ Generic AI = table stakes. Customized AI = defensible moat.

7. Google is about to rewire its monetization model

Don’t sleep on Gemini.

It’s already a strong product—and if you’ve only used ChatGPT, you’re missing part of the map.

We see Google starting to connect the dots between AI and its core business: ads. Expect more AI-native ad formats and monetization hooks within Gemini and Search.

→ Google won’t give up revenue. It will repackage it.