Operator’s Notes

Approach

I’ve run the ads, and I’ve paid the bills.

I spent five years running Amazon PPC on the agency side, then built and ran my own brand with my own money. The first taught me how accounts are managed. The second taught me what the monthly report leaves out.

Konstantin, founder of Operator’s Notes
Konstantin, Operator’s Notes

Decide on evidence. Get it done. Measure what moved. Then decide better.

On the agency side, the job is the ad account, and the scoreboard is usually ACOS. That’s a reasonable job. It just isn’t the owner’s job. The owner lives on total sales, organic rank, share and margin, and those rarely appear in the same report.

Running my own brand made that concrete. I once switched suppliers to save a few dollars a unit without pricing the downside. Quality slipped, reviews sank, conversion followed, and no ad metric warned me. It was an expensive lesson in reading the business rather than the dashboard.

So the work starts with one question: which explanation does your account actually support? Then it ranks the few decisions worth making, gets the agreed ones implemented, and checks what happened against a baseline written down in advance.

Principles

Six rules I hold the work to.

They are the home-page beliefs, written as commitments. Each one is something you can hold the work to.

  • Check the explanation before building the fix.

    Every stalled number arrives with a story and a fix attached. I test the story against the account first, including my own first hypothesis. The symptom creates the question; the evidence narrows it.

  • Every decision spends something.

    Budget, a month of waiting, the team’s attention, a ranking you put at risk. I treat each move as an allocation of money, time and attention, and rank decisions by what they’re worth, how confident we can be, and what they cost to try.

  • Say how sure you are, and what would prove you wrong.

    A finding without its evidence is an opinion. Each one comes with a confidence level and a falsifier. If your agency’s explanation holds up, I’ll tell you that too.

  • Leave alone what’s working.

    Half the value of a diagnosis can be the list of things nobody should touch yet. Changing a working account structure just to look busy is a cost, not a contribution.

  • Own the implementation, or hand it over cleanly.

    A plan that becomes the owner’s project isn’t finished. Where it makes sense I take responsibility for getting the agreed work done; where it doesn’t, the findings are written so your team can run them.

  • Write the result down, especially when the prediction was wrong.

    Before a change ships, the baseline and the expected result are recorded. Afterwards: held, partly held, or wrong, and why. That record is what makes the next decision better.

AI does the legwork. A person makes the call.

I use AI and automation backstage: pulling and reconciling data, surfacing patterns, drafting. It means more of my time goes to judgment and less to pulling reports.

It doesn’t decide anything. Which explanation holds, what goes first, what to leave alone, and anything that moves money or changes a listing is a human judgment, and it goes through your approval.

Background.

Agency years
Five years of agency-side Amazon PPC, strategy and hands-on execution. Managed $1M+ a month in Amazon ad spend in the Thermal Clothing category. Ran Amazon PPC growth for a category-leading D2C rug brand.
Own brand
Built an FBA brand end to end with my own capital: sourced, launched, operated. Later added a D2C store on Shopify, with Meta and TikTok.
Now
Operator’s Notes: Amazon Growth Diagnostic, Rebuild and Partner for established brands, plus public case files for anyone who wants to learn the method.

Own a brand you’ve grown tired of running? I’m buying. → Get in touch

See the method on a real account first.

Case File 01 shows the checks, the evidence and the three decisions, start to finish. If it matches what you’re living through, the Diagnostic is the same method on your account.