How to Choose the Next Business Workflow to Automate With AI

Key takeaways

• Your first AI success is the best clue to your next one
• Favour tasks that reuse information you have already organised
• Score candidates on value, frequency, checkability and risk
• Automate the high-value step, not the whole process at once
• Every pilot needs a success measure and an end date

This guide supports our main article, Anthropic S-1: What It Reveals About Enterprise AI Adoption. The filing shows that even the biggest companies use AI in only a few teams at first, and that growth comes from finding the next worthwhile job. Here we answer one practical question for a small business: once one AI task is working, what should you automate next?

Why does choosing the next workflow matter?

Because the second project decides whether AI becomes part of how you work or stays a side experiment. Nate B Jones points out that real workflows tend to be messy, oddly set up and poorly documented, so picking the wrong one can waste weeks and put you off AI altogether.

Picking well has a second benefit: each project you finish leaves behind connected tools, tidy information and tested instructions that make the next one quicker.

Step 1: What is already working?

Write down the AI task that is already saving you time and why it works. For example, drafting replies to common customer emails works because the questions repeat, the answers live in your FAQ and you check every draft before it goes.

Note three things about it:

• What information it uses (your FAQ, price list, past emails)
• Who checks the output and how long that takes
• What result it improves (faster replies, fewer missed enquiries)

Step 2: How do you list candidate workflows?

List tasks that sit next to your working task or reuse the same information. These are usually the easiest wins, because the groundwork is already done. Ask yourself:

• Which jobs use the same documents or data?
• What happens just before or just after the task AI already does?
• Which weekly tasks do I put off because they are dull?
• Where do customers or suppliers wait on me?

Aim for five to ten candidates. Do not judge them yet; just get them on paper.

Step 3: How do you score each workflow?

Give each candidate a score from 1 to 5 on five measures, then add them up. Flip the risk score so that lower risk scores higher.

This scoring grid is our own suggested method, not a formula from the video.

Step 4: Where is the high-value step in the workflow?

Break the winning workflow into steps and find the one where AI adds the most value. Nate stresses aiming AI at the high-leverage point in a workflow rather than trying to automate everything.

For a weekly client report, the steps might be: gather numbers, spot what changed, write the summary, format it and send it. The high-value step is usually spotting what changed and writing the summary. Gathering numbers might already be automated, and sending should stay with you.

Step 5: How do you run a small pilot?

Run the new AI step for two weeks with one clear measure of success.

1. Write the measure first, for example “report takes under 20 minutes, with no more than one correction”
2. Record a baseline: how long it takes today without AI
3. Give the AI the same information each time and save your instructions so runs are consistent
4. Check every output during the pilot and note what you corrected
5. Set an end date in your calendar to review

Step 6: How do you decide to keep, change or drop it?

At the end date, compare results with your baseline and choose one of three options:

• Keep: it met the measure, so write down the instructions and make it part of your routine
• Change: it nearly worked, so fix the instructions or the information and run one more short pilot
• Drop: it did not save time or needed too much checking, so stop and move to the next candidate

Then go back to your scored list. Dropping a pilot is useful learning, not failure.

What are good next workflows for a small business?

These are example ideas, not results from a specific business.

Frequently asked questions

The bottom line

The best next AI workflow is usually the one next to what already works. List the nearby tasks, score them honestly, aim AI at the most valuable step and test it for two weeks. Your next step: pick five candidates today.

Related posts

• Anthropic S-1: What It Reveals About Enterprise AI Adoption
• How to Plan for AI Automation in a Small Team Without Panic
• How to Measure the ROI of an AI Tool in a Small Business

Garry Knight

I'm Garry Knight, the person behind Prodify Digital. I write about email list building, email marketing, SEO, AI search and the tools that connect them. My aim is to make online marketing easier to understand, so creators and small business owners can make informed decisions about building an audience and keeping people engaged. Here you'll find straightforward guides and product reviews that explain what something does, where it fits and which limitations matter. The focus is on clear explanations and useful next steps—not hype, shortcuts or promises of easy earnings.

See Full Bio