How our copy team uses AI: why good habits beat good prompts
The biggest change artificial intelligence (AI) brought to our copy team isn’t that it helps with writing. It’s that it helps us work more intentionally before writing begins.
We expected AI to help with drafts, research and the routine knowledge work that takes up a writer’s day. It has. But the more meaningful shift came earlier: in how the team tests ideas, builds a process and makes decisions before a single word is written.
This is the third piece in our series on where AI actually stands. The first two covered what clients want and how we’re approaching AI as a full-service B2B creative agency. This one hands the mic to the people doing the work.
From three steps to seven
The old copy workflow was simple: research, draft, edit. Three steps, mostly linear.
Today, our copy team uses seven steps: research, strategy, outline, phased drafting, AI review, human edit and adaptation. The process is longer on purpose. Each step earns its place because skipping one creates problems later — from a weak strategic foundation to a draft that needs more rework than it should.
The extra steps aren’t overhead. They’re where judgment happens.
Good habits matter more than good prompts
Ask most people what’s holding back their AI use, and they’ll say they need better prompts. Our copy team found something different: how you work with AI matters more than the exact words you use.
Three habits changed our workflow. We ask AI what it needs before starting, build an outline before drafting and invite pushback, then critique the result against logic, audience alignment, format and polish.
These are small changes in how we start a conversation with AI, but they can make a meaningful difference in what comes back.
Where the value shows up
Our effort used to concentrate on production. Writing the draft took more time than anything else, with editing, research, strategy and process around it.
That balance has flipped. Strategy now takes the largest share of our time, followed by editorial judgment and process design. Research and analysis have moved up. Production, the part AI now handles more of, has dropped to the smallest share.
That doesn’t mean we’re doing less. It means we’re spending more time on the parts of the work that require human judgment.
What doesn’t change
AI is useful for first drafts, research, organizing information, summarizing and brainstorming. It does not replace our judgment, fact-checking, understanding of a specific client’s context or strategic decision-making. Those responsibilities stay with our team every time, before anything reaches a client.
For clients, the value of AI isn’t about shortcutting the work. It’s about strengthening the thinking behind it.
Used well, AI gives our team more room to test ideas, pressure-check assumptions and bring sharper strategic judgment to the work.
What’s next for the team
Today, our team’s AI use is strong. Knowledge still lives with individuals, and the opportunity is to make more of that knowledge shared, repeatable and easier to scale. We have solid client context, mature workflows and an instruction library that’s still taking shape. The next step is to make those practices easier to share across the team through a shared instruction library, reusable prompt patterns and consistent quality review. That way, the habits one person builds can strengthen everyone’s work.
This is the third in our series on where AI actually stands. Next, we hand it to our creative team, where AI plays a different role: early-stage collaborator, not finisher.
This article was originally published on LinkedIn on September 1, 2026.
