Who writes this
I’m Abhi Yadav - 3x entrepreneur, MIT alum, founder/CEO of iCustomer: a decision intelligence company building self-improving audience intelligence and measurement for humans and AI agents starting with commerce media.
Two decades building decision-driven growth engines at Google, Palo Alto Networks, American Express, Target, GM, and Nike. Early CDP pioneer helped build the category, scaled it, saw its limits firsthand.
Now I’m focused on the missing layer in an agentic, AI-first world: the one between your data stack and actual outcomes. The hard part was never the data. It’s activating decisions, not datasets.
What this is
MAD Tech (Marketing, Advertising & Data) is converging fast. Most orgs are stuck in the middle — fragmented stacks, misaligned teams, dashboards that can’t tell you what to do next. And AI automation without a decision layer makes it worse: for your CX, your brand, your budget.
Decision Loop is a biweekly newsletter on the shift from data hoarding to decision loops that compound where your stack stops reporting what happened and starts driving what’s next. Each issue:
Decision-led frameworks — practical, built from real engagements with growth teams and D2C brands
The MAD convergence — and what it means for your stack, team, and budget
Agentic AI in practice — what works in production vs. expensive chaos
Data cloud/warehouse-native strategy — the decision engine you already own but aren’t using
Retail media & new attention surfaces — where commerce media and AI search fit, and how to measure what works
Who it’s for
Whoever owns growth outcomes - CMO, CDO, CIO, VP Marketing Ops, or a title that didn’t exist two years ago. GTM and RevOps operators. Founders in MAD Tech. Data leaders bridging insights and activation. Anyone tired of buying tools to solve a decision problem.
What you won’t find
Hype. Jargon salads. Vendor press releases. LinkedIn-bait hot takes. I write from the operator seat what I’ve built, learnt from outlier brands and leaders, what I’ve watched fail, what I’m building now.
Subscribe
From data-rich and decision-poor to decision loops that compound. Free subscribers get the biweekly essay.
Have a Audience or decision intelligence challenge? Building in this space? Reply to any issue - I read everything.
