AI in the workflow · Clorox
Three days to two hours
Every month, two people spent three days in Excel stitching Circana
reports together and hunting for why revenue moved. Slow, ad hoc and
rebuilt from scratch. I consolidated every report into a single pull,
then built an AI topline analysis skill that reads the drivers behind
the number, not just the number. Same monthly question. New runtime:
about two hours. AI didn’t replace anyone’s judgment. It removed
the grind, so the team spent its time on the only thing that mattered:
deciding where to act and how to win.
Go to market · J.P. Morgan
One handbook, four buyer types
A newly formed credit group at J.P. Morgan was sitting on inventory
nobody had a sales motion for: synthetic credit, tranched exposure,
assets the bank wanted off its balance sheet. No single buyer wanted
all of it. So I stopped pitching and started segmenting. Public
finance buyers could read the tax angle. Credit analyst shops could
underwrite the risk. Hedge funds wanted the yield. Instead of a pitch
per buyer, I built one handbook organized by asset, so a client on the
West Coast roadshow could pick what was worth a second conversation.
Roadshow follow ups generated an additional $300MM in traded
notional, 200% above expectations.
Growth · Clorox
Innovation launch stack
Specialty brands were crowding the litter aisle, and Fresh Step needed
an innovation launch that would actually move. I built the whole stack
behind it: the claim, the creative, the consumer test that proved the
claim landed, and the forecast that told Walmart and Target what to
expect. The forecast and the consumer insight carried the buyer
meetings, building conviction that the innovation deserved space on
shelf. Two months after launch, sales were up 60% against the prior
period. The retailers got the win I promised them, and Fresh Step
took back the aisle.
Ecommerce · P&G
The keywords nobody was buying
ZzzQuil Kids was telling parents it worked faster. Parents weren’t
buying speed. They were buying a quiet house and a kid who slept
through the night. I moved the brand from the functional claim to the
emotional one, and the imagery from kids alone to parents and kids
together.
Then I looked at where the money was going on Amazon: paid
search sat on branded terms and a narrow phrase like “children’s
melatonin,” while the traffic lived in “kids sleep” and Spanish
category terms nobody had bought. I rebuilt the keyword list and the
listing’s hidden search terms around the language shoppers actually
used. Conversion rose 15 percent.
