Reservation economics.

A few months ago I published a note about why no-show prediction is the hardest interesting problem in hospitality. Small data, dirty labels, asymmetric costs, political interventions, delayed feedback, misaligned platform incentives. Six fights, not one.

That essay diagnosed the problem. This one is about the bigger picture I was holding in the back of my head the whole time I was writing it.

No-shows aren't a feature gap. They're a category gap.

Hospitality has a hundred software categories. POS, PMS, CRM, reservation systems, loyalty platforms, review management, contactless ordering, kitchen display systems, labor scheduling, inventory management, dynamic pricing. Each one exists because at some point, the industry decided that thing was important enough to need its own software.

What hospitality doesn't have, yet, is a category for the economics of the reservation itself. Which is strange when you think about it, because the reservation is where most of the revenue starts and where a meaningful share of the revenue ends up dying.

I want to name this gap and propose a category for it. I think it should be called reservation economics, and I think it's the most underbuilt area in hospitality software right now.

---

What reservation economics actually is

The reservation is a small, fragile contract. A guest commits to showing up. The operator commits to holding the seat. Both parties expect the other to honor the agreement. Most of the time they do. Some meaningful percentage of the time, they don't. And when they don't, the consequences ripple in both directions.

Reservation economics is the discipline of treating that contract as a real economic object worth measuring, predicting, securing, and learning from. It sits across at least three operational layers:

1. The prediction layer. Which reservations are likely to honor the contract? Which are likely to break it? This is where I've put Coverly.
2. The intervention layer. Given the prediction, what do you do about it? Smart deposit collection, friction-tuned confirmation flows, dynamic policies. This is the work most no-show prediction systems leave for the operator to figure out alone. It's also where most of them die.
3. The monitoring layer. Day to day, week to week, what's actually happening at the reservation level? Which patterns are anomalous? Which servers, channels, dayparts, party sizes are silently bleeding revenue?

Right now, each of these three layers is partially served by parts of existing reservation platforms, partially served by spreadsheet hacks, and mostly not served at all. There's no integrated stack. There's not even an integrated language for talking about it.

The reservation platforms (OpenTable, Resy, Tock, SevenRooms) could build any of these layers. As I argued in the no-show essay, they don't, because reducing no-shows doesn't directly improve their revenue. The platform sells reservation volume and seat management. The operator absorbs the cost of broken reservations. Misalignment of incentives at a structural level.

So the category sits open.

Why the existing categories don't cover it

The most common response when I describe this to operators is: "isn't this just CRM?" Or: "isn't this just reservation system features?" Worth answering both honestly.

CRM is about the guest as a long-term relationship. Lifetime value, segmentation, win-back campaigns, loyalty programs. CRM cares whether a guest comes back over six months. Reservation economics cares whether this specific reservation tonight honors the contract. Different time horizon, different unit of analysis, different decisions.

Reservation systems are about logistics. Show me the calendar. Open the table. Confirm the booking. Send the reminder. These platforms care about fulfillment of the reservation as a workflow. They don't care about the economics of whether it will honor itself. Crucially, they have no incentive to care.

Revenue management in restaurants is mostly demand-side: pricing, surge, daypart strategy. It doesn't model the supply-side leakage of broken reservations.

Operations dashboards report on what happened. They don't predict, don't intervene, and rarely connect the daily pattern back to the underlying economic contract.

Reservation economics sits in the negative space between all of these. It's the layer that asks: given that the reservation is a small economic contract, how do we make it more reliable, more predictable, and more accountable for both sides?

That layer is missing. Not buried. Not partial. Missing.

The shape of a real reservation economics stack

If someone were going to build a serious software practice in this category (and I am) there are three products that have to exist for the category to make sense. Each one is standalone valuable. Together they're a system.

Coverly: predict. Score every reservation before it happens. Use the data hospitality already has (party size, lead time, channel, history, weather, context) to estimate the probability the contract will be honored. Output: a risk score, plus the reasons behind it. The operator now knows which reservations to worry about.

Earnest: secure. Convert Coverly's prediction into action. For high-risk reservations, request a small refundable deposit with friction tuned to the operator's brand. For medium-risk, send a smarter confirmation flow. For low-risk, do nothing. The operator sets the policy once. Earnest enforces it quietly. Output: the contract is more reliable, the deposit either holds the guest or covers the seat.

Posted: watch. Read the daily reservation, POS, and Coverly data in one place. Surface what's anomalous, what changed, what to watch for tonight. Posted is the eyes on the floor when the operator can't be there in person. Output: problems caught earlier, decisions made with better information, patterns visible that were previously hidden.

Three products. One operator. One database. One coherent story.

The names matter. Coverly is the prediction layer because covers are what's being protected. Earnest is the deposit layer because earnest money is a real, understood economic instrument. Posted is the monitoring layer because posted-up is how restaurant operators describe vigilant presence on the floor. The vocabulary is restaurant-native. None of these names sound like SaaS.

Why this isn't a startup pitch

I want to be careful about something. This essay names a category and proposes three products. That sounds like a startup pitch, and I am not running a startup.

Cornerstone is a personal studio. The work happens on nights and weekends, in service of a longer practice rather than a funding event. I'm not raising capital. I'm not hiring a team. I'm not optimizing for a market exit.

What I'm doing is building three products I think should exist, in a category I think should exist, because I've spent twelve years inside hospitality and watched this gap stay open. Operators have lived with it. Platforms have ignored it. Consultants have monetized around it. Nobody has built the thing that should be there.

The honest reason this category hasn't been built is that it requires a builder with operator empathy and the patience to deal with hospitality's six structural problems (small data, dirty labels, asymmetric costs, political interventions, delayed feedback, misaligned platform incentives). The platforms have the data but not the incentive. The startups have the incentive but not the empathy. Operators have the empathy but not the builder skills.

A personal studio working from operator empathy, on a slow timeline, with no obligation to anyone except the work, is a strange but possibly correct shape for this category. We'll see.

What I'm not claiming

A few honest caveats.

I am not claiming reservation economics will save restaurants. The structural challenges in independent restaurant operations are vast and mostly not software-shaped. What I'm claiming is that within the part of those challenges that is software-shaped, this is a gap worth filling.

I am not claiming the three products I named are the only three. There could be a fourth and a fifth. There could be products that bend the category in directions I haven't thought about. Categories grow. This is the version that makes sense to me now.

I am not claiming I'll definitely ship all three. Coverly is in motion. Earnest is named and in the workshop. Posted is named and in the workshop. The ambition is the full stack. The discipline is one product at a time.

And I am not claiming this is the most important problem in hospitality. It's the most interesting one to me, given my background, given my taste, given what I can actually contribute. Other people will work on labor, food cost, supply chain, training. Good. The reservation economics gap is the one I can't stop thinking about. So I'm building it.

What this means for operators reading this

If you're an operator and you've recognized your own experience in any of this (the Saturday night you held a six-top for forty-five minutes, the deposit policy you can't quite bring yourself to implement, the cover count on Tuesday that looks weird but you can't figure out why) I'd love to talk.

The studio is small. Coverly is the product furthest along. Earnest and Posted are real bets that will take real time to build. I'm building them in the open, slowly, with the operators who care enough to talk to me about the problems they actually have.

This essay is the thesis statement. The work is the proof.