A few weeks ago, I was reading OpenAI’s post on the introduction of agents for the workplace when one detail stood out. They described the state of AI adoption for most businesses as “dispersed notifications and partially structured workflows” collected over two years. That’s a real test from the company that sparked the AI boom in the first place.
Despite the headlines, many organizations still haven’t fundamentally changed the way work is done. AI can generate solid content in less than a minute, but fast jobs don’t automatically build fast organizations.
Most marketing teams have spent the last two years doing exactly what they were encouraged to do. Each professional figured out how to make AI useful within their workflow.
A content specialist writes newsletter snippets on ChatGPT. The designer produces images that comply with the product in Firefly. An email marketer built a QA workflow that saves hours every week. Administrators use ChatGPT to check copy before it is sent.
That is real progress. But also unconnected progress.
Your customers are searching everywhere. Make sure it’s your product he appears.
The SEO toolkit you know, and the AI visibility data you need.
Start a Free Trial
Start with
The workflow bottleneck AI didn’t fix
In my upcoming book, “Hyperadaptive: Rewiring the Enterprise to Become AI-Native,” I go over an integrated marketing team I call Meridian Digital. They were tasked with turning a Monday morning blog post into a Tuesday email newsletter. In the pre-AI world, that task took four days.
With each person using AI in their work, each person’s jobs became faster, but a newspaper still takes four days to produce. Handoffs are still human. Waiting among people is still people. Permits are still people. Everyone immediately got 30% for each of their jobs, and the overall process didn’t change.
That process is how local development looks at the organizational scale. This pattern isn’t unique to marketing – any professional team handling individual AI is likely to sit somewhere close to this workflow.
When I share this example with marketing leaders, most say something along the lines of, “That’s almost exactly where we are.”
How to connect specials together
The mistake I see organizations making is thinking that this bottle is a tool problem. That’s not the case. Workplace agents, a platform for marketing agents by Jasper, Copilot, and Claude Skills. All of this is a logical progression of the platform level, and it all puts you somewhere where you can do good work.
But the special linking function is not a feature you can buy. It is a structural change, which is why less than 90% of the successful 10% see real impact with AI.
When was the last time your team took an honest look at how much of your AI progress is staying within the niche rather than across the board?
I’ve been mapping the journey organizations take from the initial confusion of AI to native AI. It has six waypoints, and the most important part of this conversation is the middle three.
- Classification of AI: Power users pull ahead of everyone else. The most powerful use cases come from packaging. Progress is high because the whole team is still figuring out what AI means for their role.
- Area continuity: Individual automations are starting to appear – one for content, one for design, one for reporting – but they are closed. There is no repeatable way to measure them across the group.
- Coordinated progress: A network of AI open source forms. Default is connecting. An organization starts to behave like an AI system rather than a collection of AI-enabled people.
Most of the sales teams I work with sit on a balance between the Bifurcation and the Landscape progression. They are not currently in Integrated development because no one has been asked to own the work of connecting professionals.
You can’t buy your way through that gap. You have to build it. That role (I call it AI leadership), pattern (opening hub), and cadence that transforms successful workflows into shared infrastructure, updates, retires what doesn’t work, and improves what works.
Meridian Digital’s newsletter workflow looks very different when it comes to integrated progress. In that case, the system finds new blog posts, draws variations of the three newsletters, generates images that comply with the brand, checks the quality, and presents options to the manager during his general content review.
Lead time is from 4 days to 1 day. Cycle time drops from two hours to one. Jobs are flexible. A content specialist becomes a content system programmer. The designer is shifting to visual direction and monitoring areas. An email marketer is starting to build their own automations.
Your next move
I leave you with this challenge. Look at the map honestly. Many of the marketing leaders I work with find that they are a full stage behind where they thought they were, and that discovery is what makes the right next move obvious.
