Hilton + Hyatt + Lightspeed: the 'C-shaped economy' and a quiet $81M restaurant-tech divestiture

I left the desk at 4:10 Thursday afternoon with three tabs still open in the browser, and the read across them is the column. Monday it was Hilton’s Q1 print — RevPAR up 3.6%, Chris Nassetta on the call telling sell-side that the U.S. macro is shaping into a “C-shaped” economy and that the chain’s customer mix is sitting on the right end of the curve. Tuesday afternoon, a release I almost scrolled past: Lightspeed Commerce divesting its Upserve U.S. hospitality product line to Skyview Equity for up to $81M. This morning, before I’d finished the second coffee, Hyatt’s Q1 crossed at 7:01 ET — RevPAR up 5.4% globally, EPS $0.63 against a $0.57 consensus, full-year adjusted EBITDA growth reaffirmed at 13-18%.

Three prints. Two giant branded-hotel operators compounding. One mid-cap restaurant POS company quietly selling off a U.S. hospitality leg at what looks, on the math, like a forced multiple. The contrarian read I want to lay down before the news desks file these as three unrelated stories: hospitality is bifurcating, and the cut isn’t where most people think it is. It isn’t luxury vs. select-service. It isn’t branded vs. independent. The cut is chain-tech bundles vs. point solutions — and the prints this week say bundles are winning, point solutions are getting carved off, and the valuation gap is opening fast.

The Hilton + Hyatt read: branded-bundle leverage is showing

Start with the two hotel numbers, because they’re closer than the headline narrative makes them look. Hilton printed RevPAR +3.6% on Monday; Hyatt printed +5.4% this morning. The Hyatt beat is partly mix — Hyatt skews more international and more luxury than Hilton’s blended footprint, and both of those buckets are running hotter than the U.S. select-service midweek base. But the directional read is the same: branded-network RevPAR is positive, comp-store occupancy is holding, and the rate side of the equation is still doing work in the second year after the post-pandemic ADR step-up.

Hyatt’s revenue line came in at $1.75B, up 1.75% year over year. That number is a tell — top-line growth is sub-2% but EBITDA growth is guided at 13-18%, which means the leverage in the model isn’t coming from RevPAR. It’s coming from the asset-light shift, from fee economics on the System of Hyatt rollout, and from the loyalty-program contribution that flows through at a much higher incremental margin than owned-hotel revenue. The EPS beat at $0.63 against the $0.57 print is a six-cent overshoot in a quarter where the macro tape is supposed to be wobbling. Six cents on a $0.57 base is not a rounding error. It’s the model finding operating leverage in a slower top-line environment, which is precisely the regime asset-light hotel companies are built for.

Hilton’s number tells the same story from a different seat. Nassetta’s “C-shaped” framing on the Q1 call was the most useful piece of macro language I’ve heard from a hotel CEO this cycle — a deliberate refusal to call the economy K-shaped, U-shaped, or recessionary, and instead to mark that the customer base Hilton serves is “curling back” toward the high-income, high-status-tier end of the loyalty curve. The companion read on the operator side, which Sarah filed Monday afternoon, is that Hilton’s domestic business-transient mix is the strongest it has been in eight quarters, and the group pace into Q3 and Q4 is sitting at low-single-digit-positive against a comp that was already tough. That isn’t a chain limping into a soft-landing scenario. That’s a chain pricing into one.

The forward read on both names is the same: branded-bundle hotels compound because the bundle has three legs reinforcing each other — distribution (the central reservation system and the OTA negotiation leverage), loyalty (the points-program economics and the cost-of-acquisition arbitrage against Booking and Expedia), and technology (the property-management stack, the AI front-desk overlay, and the integrated POS that runs F&B at scale). Each leg gets cheaper per unit as the network gets denser. Each leg gets harder to replicate for an independent operator running point solutions stitched together with middleware. The work Marriott has been doing on AI front-desk deployment is forthcoming, and Hilton’s own AI Planner audit from early March marked the same direction of travel — branded networks moving the tech stack inside the bundle rather than sourcing it from third-party SaaS vendors.

The Lightspeed carve-out: $81M says the quiet part loud

Now read the third tab. Lightspeed Commerce, Canadian mid-cap, sold its Upserve U.S. hospitality product line to Skyview Equity on Tuesday for up to $81M — $44M fixed at close, $37M in earnout tied to performance benchmarks. The line being divested supports roughly 3,200 customer locations, runs approximately $140M in FY26 revenue, generates approximately $26M in FY26 gross profit, and processes approximately $5B in gross transaction value annually. Lightspeed framed the transaction as a “strategic refocus” on its core retail and global hospitality platform, reiterated its FY27 adjusted EBITDA guide of $75-95M, and noted approximately $200M remaining on the buyback authorization.

Do the math, because the math is the story. Upserve booked through to Lightspeed at a top-line of $140M and a gross profit of $26M — that’s a gross margin of roughly 18.5%, which for a SaaS-and-payments POS business is below the line you’d want to see if you were building the long-run model on the asset. Sold for up to $81M fully earned-out. The fixed component is $44M, which on $140M of revenue is 0.31x revenue. Even at the full $81M, the multiple is 0.58x revenue, or roughly 3.1x gross profit. For context: when Lightspeed acquired Upserve in 2020 it paid $430M. Six years later it’s selling the U.S. hospitality leg for less than a fifth of that, gross of the earnout.

That spread is the cross-read I keep circling. A pure-play U.S. restaurant POS asset with $5B of GTV running through it just cleared at 0.3x to 0.6x revenue. The same week, the two largest U.S. branded hotel operators are printing EBITDA-growth guides at the upper end of the cycle. The valuation gap between “POS-only point solution serving the long tail of independent restaurants” and “branded-network hospitality compounding the bundle” just got marked, in cash, by a strategic seller and a private-equity buyer who both saw the same data room.

Skyview Equity isn’t buying Upserve because the asset is broken. They’re buying it because at $44M of fixed consideration against $26M of gross profit, the entry multiple is roughly 1.7x gross profit — and there is a real operating company underneath that can probably run the asset at a higher margin than Lightspeed was willing to inside a public-company P&L that needed to defend the FY27 EBITDA guide. The earnout takes the deal up to 3.1x gross profit if the asset performs, which is a fair multiple but not a strategic one. That’s the cleanest signal a strategic seller can send: this asset is no longer accretive to the platform, and the dispositional multiple proves it.

What Nassetta means by “C-shaped”

Nassetta’s framing matters because it gives the right vocabulary for what the spread between the two columns of this week’s tape is actually saying. A C-shaped economy is one where the top of the curve and the bottom of the curve are both running, but the middle has hollowed. In hotel-operator language: luxury and upper-upscale are pricing; economy and midscale extended-stay are filling on rate-sensitive demand; the conventional select-service midscale leg — the one with the most exposure to small-business travel and rate-shopping consumers — is the slack lane.

Translate that back to hospitality tech and the pattern lines up. The bundle plays — branded hotel companies running integrated stacks — serve the top of the C. The marketplace plays — booking aggregators, OTAs, channel managers — serve the bottom. The middle, which is independent and small-chain restaurants running stitched point solutions, is where the valuation compression is showing up. Upserve’s customer book — 3,200 locations, mostly independent operators — sits squarely in that hollowed-out middle. That’s not a coincidence with the multiple at which it cleared.

The Mews PMS roll-up thesis is the forthcoming counterweight to read against this. If branded-bundle hotels are compounding and independent-operator point solutions are getting carved off, the consolidation play on the independent side is to roll the property-management stack into a single platform with enough scale to look like a bundle to the merchant base. That’s the trade Mews has been positioning for, and the Upserve carve-out is the strongest signal yet that the same compression is going to play out in restaurant POS — either through a roll-up like Toast at the upper end or through PE-led consolidation like Skyview at the long-tail end. The M&A roundup is forthcoming; the through-line is set.

The bet: who’s next on the carve-out list

Here’s where I’d put the markers. One: the public-cap restaurant POS comparables get re-rated against the Upserve clearing price inside the next six weeks. Toast trades on a different growth profile and a much larger base, but the multiple-anchor effect is real — when a strategic seller marks a peer asset at 0.3x revenue, every other POS-only specialist in the sector now has that print in the room. Two: more strategic-seller divestitures land before Q2 closes. Shift4, NCR Voyix, and Olo all have legacy hospitality lines that sit awkwardly inside larger platforms. The Lightspeed move is a template, not an outlier. Three: branded-bundle hotel operators keep compounding into a slower top-line tape — the Q2 prints in July are where the EBITDA-growth-over-RevPAR-growth spread gets the next read.

The contrarian risk on this thesis is the obvious one: if the C-shape flattens into a true U or a soft landing, the middle fills back in, the independent restaurant operator base stabilizes, and the POS-only specialists re-rate. I’d take the other side of that. The structural cost advantage of the branded bundle isn’t a cycle phenomenon. It’s a network-density phenomenon, and network density compounds whether or not the macro cooperates.

The two hotel prints are loud. The Lightspeed release is quiet. The quiet one is the one I’d index on for the next two quarters.

— Marcus edits The Bottom Line for TableTransfers. Tips: ma@tabletransfers.com.