Keith Gorney spent 26 years building supply chain at Best Buy before taking that experience to Pivotree. His take on fast delivery: the technology isn’t the hard part. Trusting your data is.
Everybody wants same-day delivery now. Nobody wants to talk about why most retailers still can’t reliably pull it off. I sat down with Keith Gorney, who leads global supply chain services at Pivotree, and he said something that stuck with me: fast delivery isn’t new. What’s new is making it affordable and scalable. Catalog companies did next-day delivery decades ago. The difference today is customers expect it as the default, not the exception, and most retail operations are built on data too shaky to promise it.
Here’s the thing. Keith didn’t start in supply chain because he loved logistics. He got dared into it at Best Buy, 26 years ago, moving over from finance. He ended up spending his whole career there, then took that playbook to Pivotree, where he now works with retailers and manufacturers trying to make fast delivery actually work instead of just advertising it.
Why Do Fast Delivery Promises Keep Failing?
Amazon didn’t set out to build same-day delivery as a headline feature. Keith’s point is that it emerged from the infrastructure they’d already built to distribute to major markets, and once they had it, they sold it as a service. That’s Amazon Prime. Everyone else has spent the years since trying to catch up without the same foundation.
And a lot of the early attempts were rough. Keith brought up Walmart’s experiment having employees deliver packages on their way home from the store, in their personal cars. Walmart wasn’t insuring those cars for deliveries, so when something went wrong, Walmart was on the hook. The idea got killed fast.
“If a customer is wanting to come to your location and you can’t tell them whether you have it or not on their phone or on the computer, they’re not going to come. Or they’re going to look for it somewhere else.”
— Keith Gorney, Head of Global Supply Chain Services, Pivotree
That’s the part people skip past. Fast delivery isn’t a shipping problem first. It’s an inventory-truth problem first.
What Data Do You Actually Need to Promise Fast Delivery?
Keith broke it down in a way I hadn’t heard framed this cleanly. You can’t promise a customer something you don’t have, so step one is knowing, in real time or close to it, what inventory actually exists at a specific location. Not warehouse-level. Store-level. Then it gets harder: you’re not just checking if one product is available, you’re checking if an entire basket is available, because a customer isn’t ordering one SKU, they’re ordering a cart.
Then comes labor. Then comes the carrier. Then comes what happens when something breaks the promise mid-fulfillment, a fraud hold, a stockout, a delay, and whether your systems can automatically expedite picking or upgrade shipping to still hit the window you promised.
“You will lose over 50% of your customers on one failed transaction. So when you start to go down this path and evaluate this, you have to get it right, which requires that resiliency of your technology and a trusted repeatable process.”
— Keith Gorney, Head of Global Supply Chain Services, Pivotree
That stat alone should change how retailers think about fulfillment. It’s not a nice-to-have. One bad delivery experience and you’ve lost the customer, full stop.
Keith also told a story that’s a perfect, dumb example of what bad data actually costs you. A Best Buy customer had to buy three different countertop microwaves because the manufacturer’s measurements were wrong twice in a row, once because they didn’t account for rubber stoppers on the bottom, once because they didn’t measure the cord bend in the back. Small data errors, real customer pain, real cost to the business.
Is the In-Store Experience Actually Going Away?
I’ve heard this prediction for twenty years and it still hasn’t happened. Keith doesn’t think it will. People want to see, touch, and talk to someone about a product, and that’s not changing. What has changed is how stores and online have to work together.
His example from Best Buy is a good one. When they first launched buy-online-pickup-in-store, it was bad. Ten to fifteen minute wait times at the counter. Store staff had no idea who the customer was or what they’d ordered. It only got better once corporate started giving stores actual credit for those online sales, instead of treating them as a separate channel competing with in-store revenue.
“I would love to see ecom P&Ls go away. You can’t treat customers as channel customers. You have to treat them as brand customers.”
— Keith Gorney, Head of Global Supply Chain Services, Pivotree
That’s a structural point, not a marketing one. If your internal incentives pit online against in-store, your fulfillment strategy will always be fighting itself.
How Is AI Actually Changing Supply Chain Right Now?
Two places, according to Keith. First, replenishment: forecasting demand, managing carrying costs, deciding what to buy and when, especially with tariffs adding new cost variables into that math. Second, and this is the part he works in day to day, is post-order fulfillment: figuring out the cheapest, fastest, and least at-risk location to fulfill from, factoring in things like regional demand differences and discounted versus full-price inventory.
He also raised something I hadn’t thought about: AI-prompted upsell inside the physical store. A loyalty customer walks in, a staff member’s handheld pulls up their order history, and the system prompts, “this person bought matching shoes with this dress, offer them the dress.” Not agentic AI shopping on the customer’s behalf. AI helping a human employee do their job better, in person.
“When AI learns that it doesn’t trust you, it forgets you. If it didn’t find Best Buy because its attribution wasn’t set up correctly, it’s going to say, I’m not going to look at Best Buy anymore.”
— Keith Gorney, Head of Global Supply Chain Services, Pivotree
That’s worth sitting with. As AI-driven shopping and search become more common, bad product data doesn’t just hurt your website conversion. It gets you quietly written out of the AI’s answer entirely.
The Bigger Picture
What I keep coming back to from this conversation is that fast delivery has almost nothing to do with how fast your carrier is. It’s about whether your data is trustworthy enough to make a promise in the first place. Retailers love to talk about last-mile innovation. Keith’s argument, and I buy it, is that the real differentiator is upstream: accurate inventory, unified basket visibility, and systems that can react in real time when something goes wrong.
If you’re a retailer still treating your online and in-store operations as separate P&Ls, or still trusting manufacturer spec sheets without verifying them, this is your sign to fix the foundation before you invest more in the last mile.
Connect with Keith Gorney on LinkedIn or learn more about Pivotree’s supply chain and commerce work at pivotree.com.
Ready to make sure your own customer and identity data is something you can actually trust? Talk to a BDEX expert about how clean identity data powers better decisions across your business.
This article was adapted from an episode of Deconstructing Data, BDEX’s weekly podcast on data-driven marketing. Tune in live every Thursday at 4:15 PM Eastern on LinkedIn.
About BDEX: For companies that need clean identity data to power their products, BDEX offers unmatched quality and execution. Visit bdex.com and click “Talk to an Expert” to get started.
Watch the full episode: How Data and Delivery Are Reshaping Retail
