Zhong Xu

The through-line

Zhong Xu has spent a quarter of a century on the same problem, and he is candid about how long that is: "I started my career building 1,000 'ugly' websites for local Chinese restaurants. 25 years later, the software I've built powers nearly 200,000 restaurants worldwide" [13]. The origin was his father's POS system for Asian restaurants and a job he was "forced" into at sixteen, but the formative decision was refusing to code from a distance: "I worked for restaurants, waitered at tables, and learned the relentless, beautiful grind of the floor from the inside out. I wanted to understand their world completely before building the technology to support it" [13]. That insistence on operational reality underpins everything else, including his flat rejection of the overnight-success story, which in his telling takes a decade and in his own case two [19].

The ambition has escalated steadily in scope while the framing has stayed constant: build the plumbing, not the storefront. Deliverect has been described in turn as an API layer connecting restaurants to digital channels so they can focus on food rather than delivery logistics [16], as "the operating backbone for digital food" [17], and most recently as the system of record for a brand's entire digital business, spanning "their own apps, kiosks, web, first-party ordering, catering, 3pd marketplaces and the customer relationships that come with it" [6]. The counts track that widening: roughly 40,000 restaurants in 40 markets [17], then 55,000 locations in 52 countries processing nearly 800 million orders in 2024 [18], 65,000 in 52 countries at around 450 million orders annually [16], 60,000 locations on a roughly $100M ARR trajectory [19], and by Q2 2026 more than 100,000 customer locations carrying over $15B of food delivery, about 7% of all food delivery outside China [6].

On being the engine, not the experience

Xu draws a hard line between infrastructure and interface, and treats it as a partnership doctrine rather than a product limitation: "we build the engine and we collaborate with best-in-class partners to give enterprise brands the custom experiences they want" [1]. The division of labour with Bounteous is his worked example, with Deliverect handling "rock-solid order routing direct to the kitchen with zero manual re-entry" while the NOMNOM™ platform delivers the guest-facing personalization, so that "restaurants get the exact customization they want without sacrificing operational speed" [1]. What he says he values in the Q2 2026 numbers is not the location count itself but "what those locations are turning us into" as brands adopt the platform as the engine behind everything digital they do [6].

The surface area keeps expanding along the same logic. Acquiring Tobasco, a European kiosk company, was framed as a natural evolution from online ordering into in-store ordering, so menus and promotions can be managed across every channel from one place [18]. The same reasoning extends past restaurants entirely, into retailers joining "the food convenience transformation" alongside customers like Spar Holding Netherlands and Q8 [2]. Being trusted as core infrastructure carries obligations he treats as first-order: SOC 2 compliance is a point of pride and a statement that "we take security serious" [12].

On why digital ordering became existential

The market shift is the fact Xu returns to most often. When the company was founded in 2017-2018, large QSR chains drew roughly 10 percent of revenue from digital channels; that figure is now close to 80 percent, which makes digital sales a survival question rather than a growth channel [16]. COVID is the hinge: it compressed ten years of restaurant digitalization into 12 to 24 months, and Deliverect, which had originally intended to stay a 20 to 30 person company, responded by raising over $250M and scaling globally [19]. The practical offer to operators is a unified system covering menus, workflows and POS integration across channels like Uber Eats, Deliveroo and Just Eat, serving independents and chains such as KFC and Burger King alike [17].

None of this is presented as pure market opportunity. "These restaurants are the glue of a society" [19], and Xu takes visible pride in individual customers, from supporting a favourite coffee brand's US entry [11] to Deliverect's position as a global McDonald's supplier, where what he admires is "their absolute dedication to perfection and hospitality" and the way it "genuinely inspires me to raise the bar for our own team" [15]. The company also feeds demand intelligence back to operators, publishing analysis of what happens to digital sales and delivery orders when an event like the World Cup hits [14].

On distribution beating product

This is his sharpest correction to founders who think like him. The biggest misconception of product-minded people is that the best product wins; in fact, "the only factor that will allow you to win is how big your distribution network is" [19]. He observes the pattern by founder generation: first-timers obsess over product, second and third-time founders over distribution [19]. Deliverect reached 50 countries not by selling directly but by partnering with 10 to 20 point-of-sale companies a month and letting them distribute the software, the same instinct that led his earlier company to use the App Store as a distribution channel for iPad POS back in 2010 [19]. The partner-first posture visible in enterprise deals today is the same principle at a different altitude [1].

He also prefers a different scoreboard to the usual one. Rather than churn, Deliverect tracks net retention rate per customer cohort in dollar value, currently above 120%, which he calls best in class [19].

On the US versus Europe

The commercial difference is one of magnitude and of buying behaviour. "In the US I signed enterprise chain, it's 5,000 locations, right, because it's over a continent instead of a country," against 250 to 500 locations for a European enterprise deal [19]. The sales motion differs too: American CEOs want to meet the founder personally even on midsize deals, while European buyers are number-driven and delegate to their teams [19].

Underneath that sits a cultural diagnosis he applies to his own country: "I think in Belgium, you know, it's very much what if we fail" [19]. His answer is a refusal of that framing, whether as "Shoot for the moon, even if you miss you land among the stars" or "Don't limit yourself in what you think you can't do. You have no idea what you actually can" [19]. He extends the same critique to Europe's position in AI, where overregulation and fragmentation are the weakness; the Paris AI Summit struck him as "very France" rather than continental, while the US centralizes capital and China doubles down massively when pushed [19].

On AI, and what it does not change

Xu's advice to engineers is blunt and unsentimental: "Well, it will take your job, so you better know how to use it and increase your productivity by two, three hundred percent" [19]. The corollary is that AI itself is commoditized, so in enterprise sales the differentiator will remain people and relationships [19]. The product expression of this is voice: a partnership with SoundHound AI to fully automate voice ordering "from the moment a customer speaks to the moment the kitchen gets the ticket," across drive-thru, phone, kiosk and in-car, with real-time menus and orders going straight to POS [10]. The problem being solved is an operational one he knows from the floor, since "the ringing phone in a busy restaurant has been breaking service rhythm for decades" [10].

On engineering culture rather than hoping for it

"People think people management is fluid. No, you can engineer it" [19]. He argues culture must be written down through handbooks, rituals and processes before a company reaches roughly 50 people, citing his previous public company, which only started at 100 and never fixed it [19]. Hiring order is mindset first, skill second: young, hungry, naive people see solutions, while seasoned 30 to 40 year olds often see only problems, and his evidence is the 25-year-olds who relocated to run Australia during lockdown [19]. He remains an active recruiter for the company, promoting open roles and the teams behind them [3][4].

His operating mechanics follow from the same engineering mindset. He favours recurring meetings in large groups, such as a weekly leadership forum, so information reaches everyone unfiltered, because messages get diluted when cascaded through VPs, and he considers recurring 1:1s often unnecessary [19]. On morale, he holds that employees are happier and burn out less when they are busy and achieving meaningful things, and that a stuck team is usually too big, so the remedy is to break it into smaller teams and recreate the winning vibe [19]. The same belief in gathering people applies externally, in his enthusiasm for DOC 2026 in Barcelona and its lineup of speakers [5][8][9].

On raising money, and when not to

His fundraising method is a discipline rather than a campaign: send monthly updates covering both good news and bad to a list of relevant investors, and after roughly 20 emails demonstrating consistent momentum, investors will come knocking; Deliverect's last round was preempted and closed in two to three weeks [19]. He pairs that with a warning about raising at all. Only raise when there is a compelling reason, because investors expect 10 to 100x returns and liquidation preferences mean founders can walk away from a large exit with nothing if they are under water [19]. His assessment of investor contribution is measured: they bring "a little bit" of value, mainly network, hiring introductions and macro sensing, but unless they have operated companies themselves it is hard for them to tell you what to do [19].

On speed, grit and founder mode

Xu does not romanticise the work. "It's bloody tough, you know, I would not recommend it to anyone. The first year, year and a half, it's blood, sweat and tears" [19]. What carries you through is persistence, stated plainly: "If you don't give up, you're going to win" [19]. In war time he advocates operating in founder mode, making decisions and moving quickly because waiting is what kills you, which is the reasoning behind his most repeated instruction: "Whatever you want to do, do it today. Don't do two weeks, don't wait for meetings, just make it happen" [19]. He applies the same standard to the company's own trajectory, noting that "strategy and vision get you only so far the real strength is executing relentlessly toward that vision" [6].

Takeaways

• Treat distribution as the deciding variable, not product quality: Deliverect reached 50 countries by signing 10 to 20 POS partners a month rather than selling direct [19].
• Write culture down before you hit 50 people; waiting until 100, as his previous public company did, means never fixing it [19].
• Raise by sending monthly investor updates with the bad news included; after roughly 20 of them, momentum does the selling, and a preempted round can close in two to three weeks [19].
• Only raise when there is a compelling reason, because liquidation preferences can leave founders with nothing even in a large exit [19].
• Measure net dollar retention by customer cohort instead of churn; above 120% is best in class [19].
• Build the engine and let partners own the guest experience, as with Deliverect handling order routing to the kitchen with zero manual re-entry while Bounteous' NOMNOM™ platform handles personalization [1].
• Digital went from about 10 percent of large QSR revenue at founding to nearly 80 percent, which reframes digital ordering as survival infrastructure [16].
• Tell engineers the truth about AI: it will take the job, so use it now to lift productivity by 200 to 300%, while relationships stay the differentiator in enterprise sales [19].
• Hire for mindset before skill, since hungry, naive people see solutions where more seasoned hires see problems [19].