23 Years Inside Microsoft's Supply Chain: A Conversation with David Warrick

Last week, Graham Parker sat down with David Warrick for the inaugural Ship Angel fireside chat. David spent 23 years at Microsoft running global supply chain, now serves as Executive Vice President, Enterprise at Overhaul, and is one of the most candid operator voices in the industry. I watched the full hour. These are the moments that stayed with me.

‍

What it looks like to build supply chain at Microsoft scale

David joined Microsoft in 1999. At the time, the company had two products: Windows and Office. He left in 2022, having helped ship every major product release in between, from the first Xbox to the latest Surface. Along the way he built distribution centers, ran factories in southern China, and led the supply chain integration of the Nokia acquisition.

‍

His most memorable story was about Halo 3. The launch required around 400,000 units to land in 60 countries on the same day. Everything went smoothly except for one outlet in Alice Springs, where a senior Microsoft executive happened to walk in and find no copies on the shelf. David got the call. "399,998 copies of all successfully got to the right place at the right time, and I get pinged on the two that aren't." His point: supply chain practitioners are known for their last mistake.

‍

But the more interesting takeaway is what Microsoft built around that level of pressure. They adopted Hau Lee's AAA Supply Chain framework and put agility at the centre of every decision. They built a team called "digital supply chain" specifically to outpace the physical one. And they pioneered what David called "in-sourced outsourcing," where Microsoft staff embedded in contract manufacturer facilities to push the ecosystem to a higher operational standard. "Outsourcing is not abdication," he said. Most companies still treat it as one.

‍

What's actually changed in 25 years (and what hasn't)

David's view on the last quarter-century of supply chain software is sober. "I thought I'd actually fixed this 25 years ago," he said, recounting a recent conversation with a company still running on AS400. The systems that anchored supply chain in the late 90s, the Manugistics, the Manhattans, and Excel, are still anchoring it now.

‍

The bigger shift is structural. For most of the last 25 years, the physical supply chain ran ahead of the digital one. That dynamic has flipped. AI and modern software are now leaping past the physical infrastructure, which means companies that fix their digital layer can get ahead without waiting on physical change.

‍

David is unsentimental about how we got here: "Disparate systems with disparate data. This is a problem that is self-created. We allowed this to happen. There was no regulation, there were no industry standards. We allowed this to happen, and we didn't pay attention to it."

‍

AI: signal versus hype

This was the section I was most curious about going in.

‍

David is bullish on AI long-term but skeptical of where the industry is right now. He pointed out that ChatGPT reached 100 million users in five weeks. The internet took 15 years to do the same. Mobile phones took eight. We've never seen technology democratize this fast.

‍

But, he argued, most of what's being sold as AI innovation in supply chain is just automation of old workflows. "We're automating the old things that we used to do, and we're calling that innovation." He referenced Amara's Law: we tend to overestimate the impact of technology in the short term and underestimate it in the long term. Right now, he thinks, we're in the trough of disillusionment.

‍

The Steve Jobs iPhone analogy stuck with me. When Jobs launched the iPhone, he framed it as three products: a widescreen iPod, a phone, and an internet connectivity device. Everyone ignored the third one, which turned out to be the actual revolution. David thinks AI in supply chain is in the same place. The real shift is still ahead. Most of the use cases that will define the next decade haven't been built yet.

‍

What David sees in the operators who get it right

The thread that ran through the whole conversation was that David has no patience for incrementalism. "I stopped everybody talking about continuous improvement," he said, "because we're not moving the dial in any direction. We either revolutionize or we do nothing. And I'm okay if we do nothing. I just wanted to be really clear that I'm not going to hide behind a cloud of ambiguity called continuous improvement."

‍

The operators he respects most aren't the ones running the biggest AI pilots or the most polished optimisation programmes. They're the ones willing to look at the real problems, admit how much of the mess is self-created, and move decisively when the moment calls for it.

‍

After watching the hour, that's the line I keep coming back to. The companies that win the next five years won't be the ones with the most ambitious AI roadmaps. They'll be the operators willing to make decisions rather than wait for a clearer signal.

‍

The full recording is here if you want the unedited hour.