Portfolio · verified account data · Jul 2024 – Apr 2026
Two supplement brands on Amazon US, 22 consecutive months. Figures below are pulled from the account monitoring report I maintained, not from a case-study deck.
Platform: Amazon Ads (Sponsored Products, Sponsored Brands, Sponsored Display). These are my own figures from the monthly account report — I have not run Google Ads at this spend level, and the application that points here says so plainly.
Account 1
Account 2
The number that matters is not the spend, it is whether efficiency degraded as it grew. It did not.
Search-term mining and negative-keyword hygiene are the same discipline on both platforms: pull the report weekly, promote what converts into its own ad group, and cut what spends without booking. Bid-to-target math is the same too — ACoS is cost per sale expressed as a percentage, and cost per booked job is the same equation with a job ticket instead of an order value.
What does not transfer is the platform surface: Google Search match types, Performance Max asset groups, Local Services Ads and Google Guaranteed, and geo/schedule bid modifiers around a dispatch board. That is the gap I am closing deliberately, and the 90-day launch plan attached to my application spells out exactly how I would structure a new Cleveland electrical account from zero.
One page: spend, calls, forms, booked jobs, cost per booked job, and revenue by campaign and by service — with the search terms I added as negatives that week and what I changed because of them. The charts above are the same format, just measuring orders instead of jobs.