The short version: James Jones spent about 30 years in technology — Dell, Microsoft, e‑commerce rebuilds, then technology distribution — before landing in food service, and he now runs digital commerce for US Foods. In this one he walks me through MOXē and Menu IQ, the tool that lets an operator photograph a handwritten recipe card and get a costed menu item back in five minutes instead of the 30 to 40 it used to take. He is candid about why the old menu profitability tool only ever reached seven or eight percent of independent restaurants, what he sees on ride‑alongs when an operator discovers he is plating a hamburger at a five dollar loss, and where this goes next: a heads‑up that avocados are going up next week, not a text the morning your case gets cut.
Guest: James Jones, Senior Vice President of Digital Commerce at US Foods, where he is responsible for the MOXē e‑commerce platform, the company’s sales CRM transformation, and most recently pricing. Products, brands and companies discussed in this episode: US Foods, MOXē, Menu IQ, US Foods Pronto, Dell, Microsoft, Amazon, Walmart.
Why has broadline distribution been so far behind the technology curve?
Because the work is old, the product is perishable, and the handoffs are endless. James Jones puts it plainly: food service companies have been around for decades — in some cases hundreds of years — and long history means accumulated ways of working that never lent themselves to automation. A company starting from scratch can blue‑sky it. A distribution network cannot.
I asked him this on purpose, and not to pick on him. I still walk into restaurants today where the guy shows up at the back door with a handwritten invoice. Retail digitized. Grocery digitized. Food service did not. Jones named three things stacked on top of the history problem.
First, immediacy. “I’m a restaurant. I’m ordering today what I anticipate to get tomorrow,” he said. “There’s not a long lead time on that.” Second, perishability — I can let shirts sit on a shelf for eight weeks, but chicken breasts will not wait. Third, the physical environment. “When you put technology in a freezer, it doesn’t do as well as when it’s sitting in a temperature‑controlled warehouse.”
And then there is the fragmentation. By Jones’s read, the three big players account for roughly 25 percent of the market at the top, and the tail drops off sharply from there. That is not a market that gets standardized by fiat.
What moved the needle, in his telling, was COVID — an accelerator that predates his time at US Foods but that everyone in the building still points to — and now AI. I would add one more from my side of the table: youth. COVID pushed a lot of long‑tenured people out, and the operators who replaced them came in from tech and other industries, looked at the back door, and said this cannot possibly be this hard.
What is US Foods actually investing in behind the truck?
Most of it never shows up on a screen the operator sees. Jones described a delivery‑window problem that is really a prediction problem: the scan data tells you a case moved, but it does not tell you what is about to happen to the truck carrying it.
“Technology can tell you when someone scanned it, but they can’t tell you that a thunderstorm’s coming, a driver takes a smoke break at this time every day, or there’s additional traffic on the road. We’re starting to use AI and predictive models to show what is the trend of this so that we can give better windows to our customers.”
James Jones, Senior Vice President of Digital Commerce, US Foods
That matters more than it sounds, because loyalty here is stubbornly human. Who is my driver? Who is my seller? Do they show up every time with everything I ordered? Jones was blunt about the cost of getting it wrong: an hour late, or a case off the back of the truck, and that operator may not serve that meal tonight. “Restaurants don’t easily forget those mistakes.”
The second investment is format. Broadline is the tradition, but US Foods is putting money into what it calls Pronto — shorter trucks running mid‑day stops — because customers increasingly want the option set they get everywhere else in their lives. If you have listened to the sensor and cold‑chain conversations we have had on the show, this is the same story from the distributor’s side of the dock; Tom Woodbury walked through what happens when nobody is watching the equipment in The $30,000 Refund Nobody Caught: MachineQ’s Tom Woodbury.
Why does a distributor’s app have to be mobile‑first?
Because that is where the customer already is. Jones cited the number that decides the design: more than 60 percent of US Foods’ independent restaurant customers use their mobile device every day. Not occasionally. Every day. So MOXē is built mobile‑first by default, with web and desktop supported rather than assumed.
I told him what I see: it used to be I call my guy and he brings what I need. Now the operator has what Dave Ramsey calls a magic wand in his pocket, and he expects to push a button and have the thing show up. Jones’s answer was that mobile‑first only works if it is also simple‑first. “We know not all of our customers are technology forward as we are. I live and breathe it every day. So we know we want to make it simple, but bring the power of AI, the power of e‑commerce to the customer.”
In practice that meant walking the whole customer experience and removing friction one step at a time. Is it easy to get set up? Is it easy to add and remove users, given the churn in a restaurant staff? And is it easy to not rebuild a shopping list from scratch every week, because the system already knows what you buy?
What is the difference between MOXē and Menu IQ?
MOXē is the platform. Menu IQ is a tool inside it. Jones drew the line clearly: MOXē is the single unified e‑commerce platform — account management, product search, shopping lists, ordering, invoices, bill pay. Inside MOXē sits a branded set of customer tools, and Menu IQ is the one where the AI lives. The IQ suffix is deliberate; it marks the tools where US Foods started embedding intelligence rather than forms.
He was pointed about why a single platform matters competitively. “A lot of our competitors have many different platforms. They may partner with someone to bring a certain technology to life, but we want to have one tool in our customers’ hands.” Operators do not have the appetite to learn fifty tools — I hear that from every multi‑unit operator, and twice as loud from independents.
How does Menu IQ build a costed menu item from a photo?
You open the app, photograph the recipe — typed or handwritten, straight out of the binder that is still sitting in most kitchens — and Menu IQ scans the ingredients and the amounts, pulls them into MOXē, and matches them against what you actually purchase. If an ingredient is missing, it looks at restaurants with a similar menu type in your area and fills in the most likely match. Then it walks you through a verification pass so you confirm the product, the ingredient and the measurement before anything is saved.
Jones says those matches usually land 80 to 90 percent accurate, which is the number that makes the whole thing viable. And the time change is the headline:
“You’ve taken something that may have taken 30 to 40 minutes in the past and it’s five minutes or less for the operator to do that.”
James Jones, Senior Vice President of Digital Commerce, US Foods
The honest part of this story is the before. US Foods had a menu profitability tool prior to Menu IQ, and Jones did not dress up how it performed. Building one menu item meant entering every ingredient, every weight, every portion size by hand, then rolling it up — and if three menu items shared an ingredient, you were editing all three. Adoption sat around seven or eight percent of independent restaurants. “I think the barrier to adoption there was that level of manual effort and setup,” he said. “Only those that are really focused on profitability are gonna go in there and do that.”
Once the item exists, you set your price, MOXē supplies the cost of what you bought, and you get theoretical profitability per item. As you draw down inventory, actual cost comes in beside it, and you can watch theoretical versus actual trend over time. That gap is the whole game, and it is the same argument the menu engineering discipline has been making for years — we went deep on it in Menu Engineering: Data‑Driven Design That Lifts Margins.
What happens when an ingredient goes out of stock on a Tuesday?
This is the question I most wanted answered, because substitutions are where every costing tool I have seen falls over. You planned around one tomato, you got a different tomato, and now your pack size, SKU, portion and cost are all wrong. That is manual cleanup on a normal shift, and nobody has time for it.
Jones’s answer has three parts. First, shopping lists in MOXē carry primary items and fallbacks, picked in advance. Second, alerts: order the ingredients for chicken parmesan and if one is short, you get notified with substitutes offered, or you search and pick your own. Third, and this is the part I liked, the alert can also route to your seller, so either of you can fix it before the order cuts.
Underneath all of it is inventory transparency, which Jones named as a genuine differentiator. “When our customers go in, they see exactly how many cases are in that particular distribution center that’s near them.” Not a green dot. The case count. That is the difference between guessing before your four o’clock cutoff and knowing.
Can you trust your distributor to calculate your food cost?
I put the objection to him straight, because it is the one every operator listening is going to raise: if the company selling me the food is also the company telling me my food cost, what stops them from moving the price to fill the gap between what I paid and the 30 percent I am trying to hit?
Jones did not flinch at it. Two answers. The first is that the tools are open. US Foods items pull in automatically because US Foods knows the price and the pack, but operators are free to enter competitors’ items, and Jones assumes they will: “We know that most restaurant operators are multi‑source. They may order from us, they may order from a competitor. So all of our tools are open.” The tools are also free — a value‑add, not a SKU, and you do not have to be exclusive to US Foods to use them.
The second answer is architectural. The target food cost percentage lives on the customer’s side and saves to the individual user who created the menu item — two chefs in the same restaurant can be running 28 percent and 32 percent side by side. “That is not something that we pull on the back end and say, they want to get a higher margin, so how do we charge more,” he said. “Our food costs come in and our pricing is set totally irrespective. It has really nothing to do with how our customer tools operate.”
He also conceded the obvious: as an operator there is no way to prove that from the outside. That is a fair answer, and I would rather have a fair answer than a polished one. If you want the fully independent version of this control, that is the case Bill Lindsey made in Restaurant Cost Management — Bill Lindsey, COGS‑Well.
What does this actually do to a restaurant’s P&L?
Jones rides along with sellers, often around Austin, Texas, where he is from. He would not name names, but the pattern was specific enough to be useful. He walks into an independent, one concept, one building. They are not doing food cost at all. The bar is profitable, they think the restaurant probably is, or they break even and that feels fine.
Then the seller sits down with them and sets the first menu items up, absorbing the setup burden the operator does not have time for. What surfaces is usually a food cost somewhere between 35 and 45 percent, and an overall business that is closer to break‑even than anyone believed.
“It’s, I can’t believe I offer this hamburger at a five dollar loss every time I plate it out there.”
James Jones, Senior Vice President of Digital Commerce, US Foods
From there it is not exotic. It is portioning — use the same amount of fries every time instead of grabbing a handful — and pricing at the right level, and watching the food cost trend instead of discovering it at year end. Jones says the operators who work that loop move from roughly four, five or six percent overall profit to around eight, which is where he sees the better performers land. That is not a rounding error on a restaurant P&L. That is new equipment, a new dish, another person on the schedule.
What I appreciated is that he named the real hurdle, which is not the software. “You’re doing back of house, you’re greeting customers, you’re making sure the truck shows up on time. And they just don’t have time to get these things set up.” Once they see it hit the bottom line, they find the time. Getting them to the first data point is the job.
Where does AI take this next for independent operators?
Forward, in the literal sense. Today MOXē can tell you the day of that an item is going out of stock. Where Jones wants it is weeks ahead of that.
“If we know that the price of avocados is gonna go up next week, next month, we wanna give our customers a heads up. Because if they’re thinking about building a new menu item and it’s going to be heavily weighted to more avocado use or eggs, and they know the price of that is going up, that may affect their timing.”
James Jones, Senior Vice President of Digital Commerce, US Foods
That is a different posture for a business that has always been day‑of and next‑day. It moves the operator from reacting to a cut case to designing a menu around a price curve. The second piece is connective: as inventory, menu profitability and point‑of‑sale data come together, Jones expects the system to message operators proactively — we see the profitability of this item eroding, take a look. US Foods is building that view for its sellers first, and will push it to customers once they trust it.
If you want the operator‑side version of that same thesis, I laid out how forward‑looking demand data changes purchasing in RTG Insights — AI‑Powered Forecasting: Cutting Food Waste Before It Starts.
Why does Amazon keep setting the bar for food distribution?
Because your customers do not compartmentalize. Jones said roughly half of US Foods’ business is small and independent restaurants, and those customers behave like consumers. They order from a grocery store today and pick it up tomorrow. Amazon puts it on the doorstep overnight. Then they walk into their own kitchen and wonder why their food supply chain cannot do the same.
“Wherever Amazon, wherever some of these players … where they go suddenly becomes the expectation of the customer,” he said. The complication is that Amazon does not carry cold chain and spoilage the way a broadline distributor does. So the question US Foods is asking itself is how to give customers the most flexible set of delivery options possible without pretending the physics are the same.
This shows up on our side of the business too. Operators look at what they can do on a phone in every other part of their life and ask why the restaurant systems cannot just do that. It is the same expectation gap Rick Rosenfield described from the guest side in California Pizza Kitchen Scaled on Culture — Rick Rosenfield.
How do you get started with MOXē and Menu IQ?
If you are already a US Foods customer, Menu IQ is already in MOXē — go to the business tools drop‑down and it is there. You can also start at usfoods.com/moxe or usfoods.com/menu-iq, both of which carry help documentation. And your seller would rather sit down and set the first items up with you than have you never start — that is exactly what Jones watches them do on ride‑alongs.
If you are not a US Foods customer, usfoods.com has a short become‑a‑customer form and someone will follow up. Worth repeating, because it is the part operators do not expect: the tools are free, and you do not have to be exclusive to use them.
Frequently asked questions
What is US Foods Menu IQ?
Menu IQ is an AI‑powered menu management and profitability tool built into US Foods’ MOXē e‑commerce platform. It scans a photographed recipe, matches ingredients to products you purchase, and returns theoretical and actual food cost per menu item along with profitability trending over time. It is included at no additional cost for US Foods customers.
How long does it take to build a menu item in Menu IQ?
Five minutes or less, according to James Jones, Senior Vice President of Digital Commerce at US Foods. The prior manual process — entering every ingredient, weight and portion size by hand — ran 30 to 40 minutes per item, which is why adoption of the older menu profitability tool sat around seven to eight percent of independent restaurants.
How accurate is Menu IQ’s recipe scanning?
Jones says the AI match is usually 80 to 90 percent accurate. Menu IQ then walks the operator through a verification step to confirm each product, ingredient and measurement before the item is saved, so the operator can correct anything the scan got wrong rather than inheriting a bad cost.
Do you have to buy exclusively from US Foods to use Menu IQ?
No. Jones was explicit that US Foods assumes most operators are multi‑source. US Foods items populate automatically because the price and pack are known, but operators can enter competitors’ items so the food cost picture reflects the whole kitchen, not just one distributor’s share of it.
What is MOXē?
MOXē is US Foods’ single unified e‑commerce platform for customers — account and user management, product search, shopping lists with primary and fallback items, ordering, invoices and bill pay, plus real‑time inventory visibility down to case counts at the customer’s local distribution center. Menu IQ and the other customer tools live inside it.
Takeaways for operators
- The barrier was never the math, it was the setup. Seven or eight percent adoption on the old tool tells you everything — operators were not refusing to cost their menu, they were refusing to spend 40 minutes an item to do it.
- Theoretical versus actual is the number that pays. If your food cost is 38 percent and your theoretical says 31, the gap is portioning and waste, and it is almost always recoverable without touching a price.
- Set your fallbacks before you need them. Primary and substitute items on your shopping list, plus alerts routed to you and your seller, turn an out‑of‑stock into a two‑minute decision instead of a lost menu item.
- Let the seller do the setup. The single highest‑leverage move is having your rep sit down and build the first ten items with you. Getting to the first data point is the hard part; maintaining it is not.
- Four to eight percent is the prize. That is the move Jones sees when operators work the loop, and on a typical independent P&L it is the difference between surviving and investing.
Listen, watch, or read the whole thing
The full, searchable transcript of this conversation is here: James Jones on US Foods, MOXē and Menu IQ — Full Transcript | RTG Ep. 353.
Thanks to our sponsors. Restaurant Technologies keeps cooking oil out of your team’s hands and off your floor — Alissa Partee walked through the closed‑loop network in Restaurant Cooking Oil, Reinvented: RTI’s Alissa Partee. And a welcome to our newest sponsor, Yelp, joining us for this season of the show.
If you have not already, please subscribe to the show wherever you listen, and drop us a note at jeremy@restauranttechnologyguys.com with who you want to hear from next. Make it a great day.
