Most owners are stuck on the wrong decision about AI. They think it’s a strategy decision. Which tools? Which vendor? Which team member owns it? What’s the budget, the roadmap, the guardrails? What happens if we pick wrong?
That’s the decision they’re trying to make. However, it’s exactly why they’re not moving.
None of that is the actual decision.
The Real First Step
The actual decision is much smaller. Which one workflow do you want to run on AI first? That’s it. That’s the whole first step. Everything else — strategy, roadmap, guardrails, budget — either follows the first workflow or doesn’t need to exist yet.
Most owners have this backwards. They think they need a strategy before they can pick a workflow. Meanwhile, they try to build that strategy, get overwhelmed, and the whole thing sits on their desk for another quarter. During that time, three competitors installed one workflow last quarter without any strategy at all. As a result, they’re already six months ahead.
What Actually Works
Here’s what I’ve learned watching this play out across every industry. The owners who get past the paralysis don’t do it by writing better strategies. Instead, they lower the stakes of the first decision until it’s no longer worth stalling on.
They pick one workflow that eats too much of someone’s time. Then, they install AI to run it. It works, or it doesn’t. If it works, they pick a second one. If it doesn’t, they’ve learned something specific about their own operations — something no strategy document would have told them. Either way, they move forward and pick the next workflow.
That’s the whole playbook. There isn’t a version that requires a six-month planning cycle. Additionally, there isn’t a version that requires the perfect tool or betting the firm on getting it right.
Why Owners Avoid Starting Small
The reason most owners don’t do this isn’t that they don’t know how. Instead, it’s that they think starting small will look unserious. They want to walk into the next partner meeting with a full AI strategy, not with “I’m going to install AI on one workflow and see how it goes.”
The full strategy sounds like leadership. Meanwhile, the one workflow sounds like tinkering. That’s the wrong instinct, and it’s costing them the year.
The Real Cost of Waiting
The full strategy takes a year to draft, get feedback on, socialize, budget, hire around, and finally launch. During that same year, the owner who installed one workflow has already added a second, then a third. Therefore, they now have a live ROI ledger showing what compounds. Meanwhile, the strategy owner is still writing the memo.
What You Should Do Instead
Here’s what I’ll give you permission to do, if you want it.
Stop Building a Strategy
You don’t need one. Instead, you need a list of your workflows and a decision about which single one goes first. Not the whole list — just the first one.
Stop Chasing the Perfect Tool
The tool doesn’t matter until the workflow is picked. Once the workflow is picked, the tool becomes obvious, or it gets picked for you.
Stop Waiting for a Point of View
Your point of view on AI forms after the first workflow, not before. Nobody has a defensible point of view on AI without having installed at least one workflow. Everyone else is just theorizing.
Stop Requiring Certainty
This decision doesn’t require certainty. It requires one workflow.
How to Pick Your First Workflow
The Assessment I run is how you pick which workflow. It’s free. You bring three workflow candidates, and we map one in full. Additionally, we baseline what it costs the firm today and what it would produce running on AI. You leave with an answer.
Book it, or don’t — that’s up to you. However, stop trying to build a strategy before you’ve installed one thing. That’s the reason most owners are still where they were a year ago.
You don’t have to figure out AI. You have to install one workflow.
You run the firm. We run the AI.
The free AI Readiness Assessment maps and baselines one of your firm’s workflows and gives you a scored recommendation. Take it here: www.impacttg.com/ai-readiness-assessment