A look inside Amazon's portfolio: How Jeff Bezos manages 30+ business models simultaneously.
Jeff Bezos started with books. Today, Amazon runs satellite networks, quantum computers, and a cloud empire that powers the internet. Thirty different business models, all at once. Sounds like chaos? It's the opposite: portfolio management as a core discipline.
From books to (almost) everything
Amazon's evolution isn't linear, it's strategic optionality: online books in 1994, AWS as an internal tool in 2002, Prime in 2005, Kindle in 2007, Alexa in 2014, Whole Foods in 2017, Pharmacy in 2020 and Kuiper satellites from 2023. Amazon never stops experimenting, not even when the core generates billions. And this isn't chaotic opportunism; it's disciplined portfolio management.
Core: the profit machines (execute phase)
Amazon.com is a $470+ billion revenue machine with razor-thin margins at enormous scale. AWS is the cash cow with roughly $90 billion revenue at 30%+ margins. Prime's 200+ million subscribers are the glue that holds the ecosystem together, and the fulfillment network is a logistics asset nobody can replicate. These four mature models generate virtually all profit, and they finance everything that's coming.
Adjacent: the growth initiatives (build/grow phase)
Alexa, Kindle, Ring, Amazon Pharmacy, Twitch and the barely visible $40+ billion advertising business are beyond the experimental phase but not yet mature. Problem-solution fit is proven; now it's about finding product-market fit and scaling. They require investment but show proof of traction.
Transformational: the moonshots (search phase)
Project Kuiper ($10+ billion satellite bet), quantum computing via Amazon Braket, Just Walk Out stores and the Zoox robotaxi are uncertain, expensive and potentially transformative. Amazon doesn't treat them as normal business units: they get different metrics, different time horizons and different governance. They're in pure search mode, testing hypotheses and learning fast.
Value spaces: how Amazon creates structure
With 30+ business models, chaos threatens. Amazon clusters its portfolio into five value spaces: Retail & Commerce, Cloud & Platform, Digital Media & Devices, Health & Wellness, and Moonshots & Emerging Tech. This serves three purposes: strategic coherence (every initiative links to a theme), resource allocation (budgets per value space), and capability reuse (initiatives share tech, data and expertise).
Four lessons from Amazon's portfolio management
One: the core finances the future; without profitable execute there's no budget for search. Two: manage phases, not projects; a search initiative doesn't get execute KPIs. Three: value spaces prevent chaos by giving direction without rigidity. Four: the portfolio is never finished; what's core today gets disrupted tomorrow, so keep evaluating, reallocating and placing new bets.
Ideas intoaction.
Explore the possibilities with Patrick van der Pijl and the BMI team. Every big shift starts with a conversation.