The AI Workflow Self-Assessment

Most AI advice starts with the tools. This starts with your time.

Before you buy anything, automate anything, or sit through another demo, you need to know where the hours actually go. Not where you think they go. Where they actually go.

This assessment takes about 20 minutes. No account required, no results emailed to you. Just a set of questions that, if you answer them honestly, will tell you where AI is worth your attention and where it isn't.

Step 1: List every recurring task you do in a week.

Not projects. Tasks. The things that happen on repeat.

Write them down. All of them. Client emails, meeting prep, follow-ups, proposals, invoices, scheduling, status updates, reports, social posts, intake forms, contract drafts. Everything that shows up more than once a week or more than a few times a month.

Don't filter yet. Just list.

Step 2: For each task, answer three questions.

How long does it take?

Write down a realistic number. Not the good weeks. The average week.

How often does it happen?

Per day, per week, per month. Pick whichever fits.

How much thinking does it actually require?

Be honest here. Is this genuinely complex judgment work, or is it mostly the same thing every time with minor variations? Most tasks fall further toward the repetitive end than people expect.

Step 3: Flag the candidates.

A task is a good AI candidate if most of the following are true:

• It happens at least a few times a week
• It follows a pattern: same structure, similar inputs, predictable outputs
• It doesn't require you specifically: a well-briefed assistant could handle a first pass
• You dread it, rush it, or push it to the end of the day
• It takes longer than it should because you're starting from scratch every time

A task is a poor AI candidate if:

• It requires live judgment calls with information you can't write down in advance
• The relationship is the product: a client hired you, not a template
• Mistakes here are expensive and hard to catch
• It happens so rarely that building a workflow costs more than just doing it

Step 4: Add up the time.

Take your flagged tasks and total the hours per week.

If the number is under two hours, AI probably isn't a priority right now. Focus on the business.

If the number is two to five hours, you have real opportunity. Pick one task and start there.

If the number is over five hours, something is eating your week that shouldn't be. That's worth fixing properly.

Step 5: Pick one.

Not three. One.

The best candidate is usually the task that is most repetitive, takes the most time, and produces the most consistent output. Proposals, intake, follow-up sequences, and meeting summaries tend to win this exercise for professional services businesses. Customer communication, estimate requests, and review responses tend to win it for local service businesses.

Pick the one that, if you got it back, would actually change how your week feels.

What to do next

If you've done this exercise and have a clear candidate, the next question is whether to build it yourself or get help. The AI Workflow Audit is the paid version of this: we do it with you, go deeper, and come out with a full action plan. But if you're ready to move on your own, the Is This Task Worth Automating? guide is the right next step.