Anthropic just published how their own marketing team uses Claude. The numbers are worth paying attention to:

Team	What they automated	Time saved
Influencer Marketing	Scripts for influencers and podcasts	100+ hours/month
Customer Marketing	Case study drafts	2.5 hours down to 30 minutes (10 hrs/week saved)
Digital Marketing	Web development workflows	5x productivity increase year-over-year
Product Marketing	Launch briefs via Skills and Projects	5-10 hours per launch
Partner Marketing	Self-serve event enablement	40% reduction in trade show prep

These aren’t hypothetical projections. They’re from Anthropic’s own team, using their own product, in production.

From Zero to Figma Plugin in 45 Minutes

The case study features Austin Lau, a growth marketer at Anthropic who had never opened a terminal before Claude Code launched. His first reaction: “I have zero idea what this product is for.”

Within weeks, he built a Figma plugin that generates ad creative variations with a single click. A process that used to take 30 minutes per batch now takes 30 seconds. He built a custom /rsa slash command that generates upload-ready Google Ads copy, cross-referenced against brand voice guidelines and performance data.

His advice: “All you need to know is how to explain your challenge in a very clear, concise manner.”

This is genuinely impressive. A non-technical marketer building functional developer tools in under an hour. The barrier between “I wish this existed” and “I just built it” has collapsed.

Now Look at What They Automated

Read that table again. Every workflow Anthropic’s marketing team automated with Claude falls into the same category: creation.

Influencer Marketing: writing scripts
Customer Marketing: drafting case studies
Digital Marketing: building web pages
Product Marketing: creating launch briefs
Partner Marketing: preparing event materials
Growth Marketing: generating ad copy variations

Notice what’s missing. No team automated:

Distributing content across channels
Scheduling campaigns based on audience behavior
Monitoring performance in real time
Adjusting messaging based on what’s working
Orchestrating cross-channel campaigns

This isn’t a criticism. It’s a design reality. Claude is a copilot. Copilots help you create. That’s what they’re built for, and Anthropic’s team proves they’re exceptional at it.

But creation is only part of the job.

Where the Draft Ends

Austin’s Figma plugin generates 50 ad variations in 30 seconds. Then someone uploads them to Google Ads, sets targeting parameters, monitors click-through rates, and adjusts bids based on performance.

The case study gets drafted in 30 minutes instead of 2.5 hours. Then someone publishes it to the blog, promotes it across social channels, emails it to the customer list, tracks engagement, and decides whether to boost distribution.

Those 100+ hours saved on influencer scripts? The scripts still need to be sent, negotiated, reviewed, scheduled, and tracked for performance across platforms.

The pattern: Claude handles the blank-page problem brilliantly. The marketer still handles everything that happens after.

The 5x Problem

Here’s the second-order effect. When Anthropic’s digital marketing team gets 5x more productive at creating web content, they don’t create the same amount. They create more. Much more.

More content means more channels to manage. More campaigns to monitor. More performance data to interpret. More cross-channel coordination. The bottleneck doesn’t disappear — it migrates downstream from creation to orchestration.

This is exactly what we described in our analysis of the Claude marketing plugin launch: the plugin makes drafting free, but free drafting creates expensive orchestration problems.

Anthropic’s own data now confirms the pattern.

Why This Is Great News

We wrote previously about the copilot-to-CMO gap — the difference between a tool that helps you do marketing and a system that does marketing. Anthropic’s case study is the most compelling proof point yet.

Not because it shows Claude’s limitations. Because it shows Claude’s success.

Every marketer who experiences the 30-minute-to-30-second transformation will never go back. They’ll draft more, create more, plan more. And then they’ll ask the inevitable question: “If AI can create all this content, why am I still manually posting, scheduling, monitoring, and optimizing it?”

That question is where copilots end and autonomous marketing begins.

Anthropic is doing extraordinary market education. They’re showing every marketer what AI can do today. The gap between what Claude creates and what happens after is where the next generation of marketing tools will live.

The Real Takeaway

Austin Lau went from never having opened a terminal to building production tools with Claude Code. That’s the copilot story — and it’s powerful.

The CMO story is what happens to all those drafts, scripts, briefs, and ad variations after they’re created. Who distributes them? Who monitors what works? Who adjusts the strategy at 2 AM when engagement patterns shift?

The copilot makes the marketer faster. The CMO makes the marketing autonomous.

If you’re already using Claude for marketing and want to see what comes after the draft, see how Lane extends the workflow →

References

How Anthropic Uses Claude in Marketing — Anthropic case study
Austin Lau: Claude Code for Marketing — Video walkthrough
Related: Claude Has a Marketing Plugin. Here’s Why That’s Good News. — Our analysis of the Cowork plugin launch
Related: Own the Workflow or Lose the Customer — What Harvey and Sierra reveal about workflow ownership
Related: The Jevons Paradox Is Coming for Marketing — Why cheaper content creation means more orchestration demand