Sysco's AI360 Hits 95% Weekly Active — A Field Operator's Read on the Sales-Tech That Actually Works

I had the Sysco Q2 FY26 call on a second laptop this morning while a contractor finished re-tiling a kitchen wall at one of the operators I’m shadowing this week. The contractor is loud. CFO Kenny Cheung was muffled. So I caught the number the third time Kevin Hourican said it.

Ninety-five percent weekly active. On AI360. Three months after I last wrote about Sysco quietly becoming a software company when Samuel covered the 90% figure from the Q1 call, the adoption number is now somewhere north of where any enterprise CRM tool I have ever rolled out as a GM lived. The exact phrasing from the transcript: “95% or more of our colleagues using the tool weekly.” Then Hourican said the line that the trade press will run with all week. “The math is very clear: if you use the tool, you sell more.”

That is not a normal earnings-call adoption number. And the framing I am going to put on it — sales-tech that actually works — is interpretation, not Sysco’s. I will defend that framing in this piece. I will also tell you where the framing breaks.

How this piece was reported

This is an operator case study built entirely from public materials. I did not interview Sysco for it. The primary source is the Q2 FY26 earnings transcript on Motley Fool, corroborated by Yahoo Finance’s earnings recap, Distribution Strategy Group’s coverage, and Digital Commerce 360’s October piece on the Q1 FY26 rollout. I cross-checked Hourican’s exact “if you use the tool, you sell more” phrasing against the transcript before reproducing it. There is no Sysco operator panel. There is no per-rep productivity number, because Sysco has not disclosed one and I refuse to invent one. Where I assert correlation versus causation, I will flag it.

That is the methodology bar for The Operator. Now the read.

What 95% adoption actually means

A 95% weekly active number on an enterprise sales tool is unusual enough that the first question to ask is what it is not measuring. It is not measuring whether AI360 is closing deals. It is not measuring whether the recommendations are good. It is not measuring whether reps are following the suggestions or quietly ignoring them while keeping the tab open on a second monitor. Weekly active is a participation metric.

So why does it matter?

Because most enterprise sales software fails at the participation step. The rep opens it twice during onboarding, hates the UX, falls back to the spreadsheet they have used since 2017, and the tool’s vendor sends a Customer Success email asking to “schedule a check-in.” I have sat through three of those rollouts as a GM and watched the same pattern each time. The denominator that 95% is calculated against is the population of Sysco’s US sales consultants — Hourican has previously described that population as “the largest sales force in foodservice,” and Sysco’s most recent filings put it in the range of ten thousand sellers. Getting 95% of ten thousand commissioned sellers to open the same tool weekly, four months after launch, is a deployment achievement on its own, separate from whether the tool’s recommendations are good.

The Q1 number was approximately 90%. Q2 is 95% or higher. Over a single quarter the adoption curve moved up another five points, on a base that was already high enough to be remarkable. That tells me three things.

First, Sysco’s reps are not being given a choice about whether to log in. Whether that pressure is explicit (manager dashboards, performance plans tied to tool engagement) or implicit (the rep’s commission math gets worse if they ignore it), the absence of an opt-out is structural. Hourican said as much, in slightly more diplomatic language, when he noted that the company is “tracking” the correlation between usage frequency and selling performance across tenure cohorts. If the company is tracking it, the reps know the company is tracking it, and the floor for “weekly active” gets very high very fast.

Second, the tool is good enough not to actively repel the people forced to use it. This is a lower bar than “actually useful,” but it is a bar a lot of enterprise software fails to clear. I am taking the absence of public revolt — no leaks, no anonymous Glassdoor surge, no union pushback I can find — as weak evidence that AI360 is not making the reps’ day measurably worse.

Third, and this is where I get cautious: adoption is not effectiveness. Hourican’s framing is causal — use the tool, sell more — but the disclosed data is correlational. Higher-frequency users outperform lower-frequency users at every tenure level, per the transcript. That is consistent with the tool causing higher sales. It is also consistent with the better sellers being more willing to try new tools, which is exactly the population you would expect to be both heavy adopters and high performers regardless of what tool you put in their hands. Sysco has not published a randomised holdout. They almost certainly will not, because running a holdout on your own salesforce when you are trying to drive adoption is corporate self-sabotage.

So when I say sales-tech that actually works, what I mean specifically is: the tool has cleared the adoption bar that most enterprise sales tools never clear, and the company has wired enough internal incentives around it that the participation number will keep climbing. Whether the underlying recommendations are causally responsible for the Q2 case-volume print is a separate empirical question that the disclosed data cannot answer. Hold that distinction.

The Hourican framing: volume correlation as the headline

Hourican’s job on these calls is to give the sell-side an explanation for the case-volume line, and on this call he gave them AI360. The narrative he constructed has three steps.

Step one: restaurant traffic is brutal. Black Box data has restaurant foot traffic down more than 200 basis points year-over-year and similarly down sequentially. That is the industry weather. Every distributor on every call this quarter is going to point at it.

Step two: Sysco’s US Foodservice local case volume grew 1.2% in Q2, a 140 basis point sequential improvement versus Q1. That is the company-specific number Hourican wants you to anchor on. Industry traffic down 200+ bps, Sysco local cases up 120 bps. The delta is the story.

Step three: AI360 is the named cause. Not the only cause — Hourican also credits the Perks 2.0 loyalty refresh, the pricing-agility initiative we covered in November, and the new-account growth motion — but the call’s rhetorical anchor is the AI tool, because it is the one that gives the analyst community a single, simple, memorable story.

This is a fine narrative. It is also a narrative under pressure. If Sysco’s Q3 print does not deliver the promised ≥2.5% local case growth that management guided to twice on the call, the AI360 story stops being a tailwind explanation and starts being an awkward question — because the adoption rate is unlikely to drop, which means the lever Hourican has been pointing to will still be at 95%+ while the result he attributed to it has gone the other way. The narrative survives a quarter of in-line performance. It does not survive a miss.

I am writing this on Tuesday morning. The Q3 call is roughly three months out. We will see.

Swap and Save: the roadmap feature operators should actually read

The most consequential thing Hourican said about AI360 on this call was not about the 95%. It was about what the tool is going to do next.

The named feature in the roadmap is “Swap and Save,” and the version Sysco is describing for the back half of FY26 is cuisine-specific. The transcript description, paraphrased: AI360 will surface prioritised product substitutions to the sales consultant during the order conversation, tailored to the cuisine the customer operates in — Italian, Mexican, Asian, American casual — with recommendations that “save the customer money” while also benefiting Sysco and the rep.

Decompose that.

For the customer (you, the operator), Swap and Save is a recommendation engine pointed at your order. The rep walks in Tuesday morning, opens the tablet, and the screen has a queue of substitution suggestions for your specific cuisine. “Swap the imported San Marzano case for our private-label San Marzano-style — same recipe outcome, $2.40 per case savings.” “Drop the 80/20 ground beef vendor for our broadline equivalent — twelve cents per pound cheaper, same protein spec.” These will often be real savings. They will also, by design, route volume toward Sysco’s private-label and house-brand SKUs, which is where the company’s margin is best. Both are true simultaneously. The recommendation is honest and commercially motivated.

For Sysco, Swap and Save is a margin tool wearing a customer-service hat. The structural insight is that the rep is now equipped with cuisine-aware substitutions in real time, rather than relying on the rep’s individual memory of the catalogue and your menu. A new rep covering your account gets the same substitution quality as a fifteen-year veteran on day one, because the model is doing the catalogue work. That collapses the productivity gap between new hires and tenured sellers — which is a separate, less-discussed reason adoption is so high. AI360 makes a new rep dangerous immediately, and it makes a veteran rep faster. Both ends of the tenure distribution win.

For the rep, Swap and Save is a commission accelerator. Private-label and house-brand SKUs tend to carry better margin, and Sysco’s incentive structure rewards that margin. The rep’s commission math improves when they pitch a successful substitution. Which is why I described the tool earlier as having structural incentives around adoption: the rep’s paycheque is downstream of the tool’s recommendations.

The operator translation is unsubtle. If you are a Sysco customer and Swap and Save ships on the timeline Hourican implied — back half of FY26, which means landing between now and June — you should expect a different sales conversation by Q4. Not necessarily a worse one. But a different one. The rep will arrive with a tablet that has done specific homework on your specific cuisine. The suggestions will be plausible. Some of them will save you real money. Some of them will be substitutions you would not have made on your own.

The discipline operators need to keep around this is the discipline operators have always needed to keep around vendor recommendations: useful, biased, both. The margin discipline we covered in the Four Margins piece gets harder when one screen owns more of your purchasing decisions. If you let the substitution stack run unchecked for two quarters, you wake up in a private-label mix you did not actively choose. That is not a Sysco-specific risk; it is a recommendation-engine risk that applies to every catalogue platform now investing in this layer. But Sysco is the first foodservice distributor to ship it at scale, and that means Sysco’s customers are the first to need an answer for it.

The answer I would give an operator I was advising: keep a quarterly substitution audit. Pull every swap your Sysco rep recommended in the previous twelve weeks. Compare the substituted SKU’s actual performance against the original — taste, plate cost, guest complaints, waste. Kill the ones that did not earn their substitution. Tell the rep which ones to stop suggesting. Treat the AI360 recommendation queue the way you treat the produce buyer’s “this week’s special” — useful inputs, not the menu plan.

The Ginsburg Foods tuck-in: small deal, large signal

Buried in the same call, Sysco confirmed it had closed a tuck-in acquisition of Ginsburg Foods, a Northeast broadline distributor. Management is guiding the deal to add approximately 50 basis points to back-half local case growth, on top of organic growth of “at least 2.1%” — which is how they get to the ≥2.5% headline number for Q3 and Q4.

This is a tiny deal by Sysco’s standards. The financial line is rounding error against the $29 billion Restaurant Depot agreement that Samuel and I have been tracking since March. But it tells you something the Restaurant Depot deal does not: Sysco is still doing the boring distributor M&A that has been the company’s growth lever for a decade, and they are doing it in regional pockets where AI360’s data advantage compounds.

Here is the underlying mechanic. When Sysco absorbs a regional broadline like Ginsburg, the acquired customer book gets onboarded into Sysco’s CRM and the sales reps inherit the AI360 tool. The model now has access to a new geography’s order patterns within weeks. The reps who came over from Ginsburg get equipped with Sysco’s recommendation stack and the productivity uplift Hourican has been pointing at. The 50 bps of case growth Sysco is forecasting from this deal is, in part, an AI360 productivity bet on a freshly inherited salesforce — the tool’s compounding advantage is that every acquisition is also a deployment.

PFG and US Foods cannot say that yet. Neither has a comparable salesforce productivity stack in market. Which is the segue I have been holding back.

The implicit threat to US Foods and Performance Food Group

Hourican did not say a single thing about US Foods or PFG on this call. He did not need to. The competitive read is implicit in the adoption number.

Sysco’s two big broadline competitors are running the same playbook Sysco was running eighteen months ago: a portfolio of digital sales tools, pricing initiatives, and customer-facing apps, each rolled out independently, each adopted in patches by reps depending on geography and manager. US Foods has its CHEF’STORE cash-and-carry footprint and its Pronto small-truck delivery model; PFG has its specialty Vistar and Performance Foodservice segments. Both have internal sales tools. Neither has, on a public earnings call, claimed a 95% weekly active number on a single integrated assistant.

That gap is the threat. If AI360 is producing 140 basis points of relative outperformance against restaurant traffic — and even if half of that is causally attributable rather than the full claimed amount — that is the kind of structural delta that compounds quarter over quarter, and it is delta the competition cannot close with a press release. Building an in-house salesforce AI assistant that 95% of your reps actually use is, as I said earlier, a deployment problem, not a model problem. Both US Foods and PFG can buy the same models Sysco is using. Neither can buy four months of in-market adoption work.

The expected counter-move is one of two things. Either the competition partners with an outside vendor — a CRM platform, a foodservice-specific AI startup, an enterprise voice-agent company — and tries to ship a comparable tool faster than building one in-house, accepting that the data advantage will be weaker because the model is not native to their order history. Or one of them buys their way in, the same way Sysco is buying its way into the cash-and-carry channel with Restaurant Depot. There is precedent for both moves in foodservice tech.

I have no scoop on either. What I have is the read that the AI360 number on this morning’s call put the rest of the broadline distributor category on a clock. Whichever of PFG or US Foods announces first, the announcement will look reactive — and the trade press will write the framing that way regardless of how long the internal project has been underway. That is the cost of letting Sysco get to 95% first.

What operators should expect next quarter

Three things, in the order they will affect your week.

The substitution conversation is going to escalate. If Swap and Save ships on the implied Q3/Q4 timeline, you will start seeing cuisine-specific substitution decks from your Sysco rep within the next two quarters. The deck quality will be noticeably higher than the generic substitution pitches your rep has been making for years. Build the quarterly substitution audit I described above before the deck arrives, not after. The audit is your defence against drifting into a private-label mix you did not choose.

Your rep’s pitch is going to feel sharper. This is a continuation of what Samuel flagged in the Q1 piece: the rep walking through your back door now has a tool that knows your order history better than your sous chef does. The version of the rep arriving in Q3 will be better at the conversation than the version arriving today, because the tool is getting better and the rep is climbing the productivity curve. Match the energy. If you have not done it recently, audit your spec sheet. Be specific with your rep about which substitutions are not on the table — proteins, signature SKUs, allergen-sensitive items. Give the tool boundaries.

Watch for the cross-channel pitch if Restaurant Depot closes. Restaurant Depot is not expected to close until well into FY27, and antitrust review could extend that further. But if it does close, the AI360 recommendation engine will get cross-channel data — your broadline orders and your cash-and-carry top-ups, in one model — and the substitution recommendations get sharper because the model finally sees your full purchasing mix. That is a 2027 problem. Bookmark it.

The falsifiable prediction

Here is what I am willing to be wrong about in public.

Hourican guided to local case growth of at least 2.5% in both Q3 and Q4 FY26. He pinned the narrative on AI360. The Q3 print lands roughly the last week of April. Three scenarios:

If Q3 local case growth comes in at 2.5% or higher, the AI360 story is the credible story Sysco told this morning and the framing of sales-tech that actually works stands. The next call will be about Swap and Save adoption, and the framing operators will need to hold becomes the substitution-discipline framing I described above.

If Q3 local case growth comes in between 1.5% and 2.5%, the story gets murkier. Adoption is still high, the tool is still doing something, but the headline number Sysco asked the sell-side to anchor on did not arrive. Expect Hourican to point at restaurant traffic getting worse — which it might — and the AI360 narrative gets softer.

If Q3 local case growth comes in below 1.5%, the framing I put on this morning’s call was wrong. Sales-tech that actually works requires the volume claim Hourican attached to the tool to land within shouting distance of his guidance. A miss that large means either the tool is not doing what the company believes it is doing, or the macro is overwhelming whatever it is doing, in which case the operator read on AI360 is that it has not yet earned the headline.

My bet, on the morning of January 27, is the middle scenario. Adoption is real, productivity uplift is real, and restaurant traffic gets a little worse before it gets better. Sysco prints something like 1.8% to 2.3% local case growth in Q3 — short of guidance, ahead of the industry, and Hourican spends the call explaining why the gap is macro and not tool. The 95% adoption number will still be there. The framing this column ran today will have to be adjusted accordingly when we cover the Q3 print.

That is the prediction. I will come back to it the last week of April.

In the meantime: the rep walks in Tuesday with a better tablet. Treat the tablet the way you treat any vendor recommendation — useful, biased, both. The audit is your job.

— Priya covers operators for The Operator. Tips: priya@tabletransfers.com.