Three weeks of every month. That’s how long one agency I’m working with spends pulling together the data, insights, story, and decks for their clients’ media reporting. Three weeks.
By the time they’re done, the month is almost over, the clients have moved on, everyone is exhausted, and they have to start the whole thing over again. Talk about a Sisyphean task.
To be clear, this is not an indictment of the agency or its people. They are unique in a lot of ways, but in this way, they are just like almost every agency we work with.
And it’s not just data collection and reporting that agencies like theirs struggle with. It’s almost everything agencies do, every day. Resource allocation, scopes of work, status updates, campaign kick-offs, financial reporting — agencies have an endless supply of repeatable tasks that consume inordinate amounts of time, energy, and, ultimately, money.
By the time the routine things are done, there’s little time left for the work that really matters. In my client’s case, there’s no time for insight development, for digging into their clients’ competitive environments, for creating reports that the client is champing at the bit to receive.
Instead, everyone moves on. Then starts the cycle again.
You’ll hardly be surprised to hear me say that AI is the solution to this problem. Or at least part of it.
But it’s not more AI tools. It’s connected AI — tools that do a lot of the heavy lifting, share context with each other, and make people’s lives easier, not harder and more fragmented.
The challenge is that this is not how most agencies have adopted AI. They add one tool at a time to solve one challenge at a time. The difference between agencies building durable advantages and the ones still pushing the boulder up the hill isn’t the tools they’re using — it’s whether those tools are connected to each other, to the agency’s knowledge, and to a shared standard for how the work gets done.
Connected AI isn’t a platform or a product. It’s three things working together. Most agencies have none of them. Some have one. The ones building durable advantages are working on all three.
Not accessible in theory — actually structured and organized so that a brief, a campaign, or a prompt can draw on it without someone manually pulling context from six different places. Brand voice. Client history. What worked and what didn’t. The institutional memory that currently lives in your best people’s heads.
For my agency client, this means competitive data, campaign benchmarks, and client context that currently lives in spreadsheets, email threads, Slack chats, and people’s heads — reassembled from scratch every single month. Put that in the system once, and the boulder stops rolling back.
There’s a meaningful difference between “we use AI sometimes” and “this is how we research, this is how we brief, this is how we review” — with AI as a defined part of each step, repeatable by anyone on the team. Not just the person who figured it out.
That agency spending three weeks on data prep isn’t failing at AI. They’re doing plenty of it. What they don’t have is AI integrated into the workflow itself — so the data is ready on day one, the insights are surfaced automatically, and the senior strategist walks in ready to think, not ready to organize.
Workflow-level integration is what turns individual experiments into organizational capability — and what turns a Sisyphean monthly grind into something that actually compounds.
Not a manifesto. Not a training program. Just: here’s how we prompt for this kind of work. Here’s what quality looks like. Here’s where a human checks the output before it goes to a client.
That last part matters most. Not because you don’t trust AI, but because your judgment is the product. It’s also your moat. Clients aren’t paying you to move fast — they’re paying you because they trust that someone who understands their business, their audience, and what’s actually at stake is accountable for what goes out the door.
A standard makes that accountability real and repeatable instead of dependent on whoever happens to be paying attention that day. It’s also what ensures that when the routine work is finally off the team’s plate, the work that replaces it is actually worth doing.
These three things — knowledge, workflow, standard — are what I’ll be going deep on at the Build a Better Agency Summit next week in Denver. But the framework works whether you’re attending or not.
Start with a conversation, not a tool. Ask two or three people on your team to walk you through how they actually used AI on something this week. Not a demo — something real. You’ll see the gaps immediately. You’ll probably also see someone doing something genuinely smart that nobody else knows about.
That’s your starting point. One workflow. Defined end to end. Two people following it the same way. That’s version one of your standard.
Try this, and I'm betting that tomorrow the rock will feel lighter and the hill less steep.