Do we have to do this every month?
A seven-figure business, a stack of printed spreadsheets, and one question from my wife. What happened when we automated the monthly data pull at Herb & Root, and where the hours went.
There’s a picture I still have in my head. Me, sitting there with a whole stack of printed Excel sheets. Thirty days of sales. Sixty days. Ninety. Looking for the seasonality.
And my wife, Julia, looks up and says, “Do we have to do this every month?”
Fair question. By then Herb & Root was a seven-figure business, and what she wanted more than anything wasn’t more revenue. It was time.
But we did have to do it. That monthly pull was how the business grew. We’d done it religiously for years. It’s where we found the areas where we were weak, and then we’d spend the hours fixing them. Skip it and you’re flying blind.
The question wasn’t whether to do the work. It was whether we had to be the ones doing it.
The map was already there.
Here’s what I noticed when I finally sat with it. Between the two of us, that ritual cost about three hours every month. And every one of those hours was predictable. Pull the same reports. Line up the same windows. Look for the same patterns. Flag the same kinds of problems.
We weren’t thinking in those three hours. We were fetching.
That’s the part people skip when they talk about automation. You can only hand off work you can describe. We could describe this one down to the step, because we’d done it by hand for years. We had a map. We knew exactly what we were looking for.
So I built the thing that does the fetching. Now the pull runs on its own, and the analysis we used to build by hand lands in our inbox, along with what we want done about it. No printouts.
And here’s the other thing: almost none of it is AI. It’s all rules. It’s deterministic, and it can be, because we know exactly what we want. Pulling the data, lining up the date ranges, comparing this month to last year: that runs the same way every time and costs almost nothing. The only place a model earns its keep is when the math gets complicated.
When you know exactly what you’re looking for, you don’t need something guessing at it. That’s how you get a number you can trust.
We didn’t rest. We moved the hours.
We started pivoting faster, because we weren’t waiting on the monthly ritual to see what was happening. The time went into advertising, into making commercials, into new channels. She got to be in more rooms. More conferences. Then we went after TikTok, and today we’re putting about $4,000 to $5,000 a month into it. That money only moves because the hours moved first.
When Amazon picked us to shoot a family commercial, part of what they saw was what we were doing with the data.
And then Herb & Root was picked for Oprah’s Favorite Things.
I’m not going to tell you the automation got us on Oprah. It didn’t. My wife and her product did that. What the automation did was give her the hours to be in the places where that kind of thing happens.
We did it backwards on purpose.
People want to start with offense. The new channel. The AI that writes the ads. The agent that runs the social account.
We played defense first. We cut out as much of the grunt work as we could, and only then did we spend the freed time going after growth. Every new push since has followed the same order. Get the data working for you first, then go do the thing only you can do.
We’re still at it. The next question on our list is how to stay on TikTok without it eating her week, and after that, user-generated content. Same approach each time. Write down what we look for, and automate the fetching.
Every business has its printed-spreadsheet ritual. The report somebody pulls every Monday. The spreadsheet that gets rebuilt every month-end. The numbers the owner checks before they’ll trust anything else.
Start with the data you already have.
Tell us the report you rebuild every month. We will map what you look for, find the number the business runs on, and show you which parts can run on their own.