North American hotels grew their guest databases by 24% in 2025 — the strongest growth of any region in the world. That’s the good news. Our 2026 Hospitality Benchmark Report tells a more complex story: 12% of profiles contain OTA-masked emails, nearly half of all profiles are missing a phone number, and 11% contain duplicate entries. Growth and health aren’t the same thing.

Chris Koehler, CMO of Twilio and a frequent business traveler, recently checked into a boutique hotel he’d visited at least ten times. The person at the front desk asked: “Have you ever stayed here before?” His reaction: “Just such an opportunity to say welcome back.” The data existed. The relationship was real. But none of it made it to the moment that mattered.

Charlie Osmond, co-founder of Triptease and a longtime student of hotel data, doesn’t see a spectrum here. He sees a binary: “Either you’ve got your data sorted and it’s really high quality and you’re going to be driving outstanding decisions, or you’ve got a data mess, and you’re just driving the wrong decisions consistently all the time.” There is no middle ground, no partial credit. And in an AI-driven hotel landscape, those gaps don’t just limit performance — they actively amplify the wrong decisions at scale.

Growth vs. health — why your database size doesn’t tell the whole story

A 24% growth rate sounds strong. And it is — more guest interactions, more potential relationships, more opportunities to drive direct bookings. But the 2026 Hospitality Benchmark Report shows that a significant portion of most North American databases is incomplete, duplicated, or unreachable, which limits the value of every campaign, call, and AI-driven decision that runs on it.

Here’s what the data shows:

• Email capture: 84%. North America leads globally with strong performance. But 16% of guest records still lack a valid email address, so they’re unreachable for email marketing.
• Phone capture: 48–58%. This is the most significant gap, and the most overlooked. With voice and SMS increasingly important channels, nearly half of all profiles are missing a phone number.
• OTA-masked emails: 12%. Guests acquired through OTA channels whose real contact information is hidden behind proxy addresses. While you have the booking, you don’t yet own the relationship.
• Merged profiles: 11%. One in nine guest profiles has been identified as a duplicate. Every duplicate is a guest being communicated with inconsistently, or not at all.

The database is the foundation for every downstream decision. Gaps in coverage and quality don’t stay contained — they grow larger across email, voice, and AI-driven personalization.

Why the AI era makes data quality exponentially more important

Data quality has always mattered, but the arrival of AI in hotel operations raises the stakes.

Osmond’s framing on the Hotel Moment podcast is worth sitting with: “I feel like in an ever-accelerating AI world, it only becomes more important. Not like a little bit more important, but like 10x, 100x, 1000x.”

AI doesn’t evaluate data quality — it works with whatever it receives. The same AI capabilities that allow hotels with clean data to make faster, more nuanced, more personalized decisions will consistently amplify wrong decisions for hotels with “dirty” data.

As Osmond puts it: “If what’s going on at the beginning is rubbish, then you’re in a really bad place.”

Koehler adds a dimension that makes the timeline more urgent. He describes a near-future where guests don’t search for hotels themselves — their personal AI agents do, drawing on rich profiles of preferences, budgets, and past stays to negotiate on their behalf. “Are you set up and ready for a world where agents are your customers on behalf of your guests?,” he asked in a recent episode of the Hotel Moment podcast.

A hotel with incomplete, fragmented guest data won’t just underperform in that world — it may not come up at all. Investing in AI on top of a data mess doesn’t close the gap between you and better-organized competitors. It widens it. The hotels that win in an AI-driven landscape will be the ones that started cleaning their data now — before amplification kicks in.

What good database health actually looks like, and where most hotels fall short

Koehler’s experience as a “road warrior” demonstrates the importance in real-world terms. He’s stayed at certain hotel brands dozens of times. He always orders extra coffee pods. He always drinks the sparkling water from the minibar. He never drinks alcohol. And yet: he runs out of coffee on every visit, the sparkling water isn’t pre-stocked, and a bottle of wine keeps appearing in the room. “Those little things matter significantly — they really aren’t that expensive but would drive such brand loyalty.” The signals are there. They’re just not making it into profiles that staff and pre-arrival teams can act on.

That’s the gap hotel database health is designed to close. Not just capturing data — activating it at the moment of impact.

Two data hygiene practices that are easy to overlook, but high-impact when you get them right:

1. Never use placeholder emails — entries like [email protected] — to fill gaps in your PMS. These corrupt your database and are nearly impossible to clean at scale. Train front desk staff to capture real contact information at check-in. A phone number or a confirmed “no email” is more valuable than a placeholder that degrades your deliverability.
2. Ensure first and last name fields are entered separately in your PMS. Revinate’s Identity Resolution depends on clean name data to accurately merge duplicate profiles. A combined name field (“John Smith” in a single entry) creates matching errors that downstream tools can’t fix.

Many database health problems are due to issues with the reservation team or front desk. The data that goes into the PMS at check-in determines the quality of every downstream decision. Getting capture right at the source is the easiest way to make sure it stays correct as you add more data to a Rich Guest Profile.

Five ways to grow and improve your hotel database starting now

Database health isn’t a one-time project — it’s an ongoing practice. But there are specific, low-friction actions that move the needle quickly.

1. Integrate your PMS with Revinate Guests. Ensure your PMS is fully connected so every guest record flows in automatically and Identity Resolution can merge duplicates in real time. This is the foundation.
2. Capture email and phone on every reservation call. Voice interactions are among the highest-quality data-capture moments available. Every call — booked or not — is an opportunity to add a verified email or phone number to a guest profile. Build this into your agent scripts and reservation SOPs.
3. Connect outlet data via Data In. Spa bookings, restaurant reservations, golf tee times — these interactions contain purchase signals that make segmentation far more precise. Connect them to Revinate Guests via the Data In API, and you’re building a more complete picture of every guest’s relationship with your property.
4. Capture cart abandonment. Website visitors who start but don’t complete a booking are high-intent leads. Cart abandonment capture feeds fresh, qualified contacts directly into your CDP — people who raised their hand and then hesitated. These are worth pursuing.
5. Run OTA conversion campaigns. Use qualification campaigns to move OTA-masked guests toward a direct relationship. These are guests with demonstrated booking intent, and the conversion data shows they respond.

You don’t need to fix everything at once. Start with PMS integration and front-desk capture — these two changes address the root cause of most database quality problems and improve everything downstream.

How to turn OTA guests into direct relationships

Just because an email is masked by an OTA doesn’t mean it’s lost. Our Identity Resolution capability merges masked OTA emails into Rich Guest Profiles when a guest books direct or provides their real contact information.

The activation tool for this segment is the OTA winback campaign. The benchmark data shows why: 52% open rate and a 2.8% conversion rate. These are not cold contacts — they’re previous guests who already know your property. The qualification campaign is the bridge between a booking you didn’t own and a relationship you do.

Osmond’s framing of where AI fits here is worth carrying: the agents — human or AI — that will win guest bookings in the future are the ones that can demonstrate they know the guest.

Koehler offers a useful lens on where this is headed. He sees travelers splitting into two segments: those who will delegate everything to AI agents, and those who will seek out the human touch as a premium in itself. Both segments reward the same thing — a hotel that knows them. For the fully “agentic” traveler, their AI needs rich profile data to match them to the right property. For the human-touch traveler, personalization at check-in is the product. Either way, the direct guest relationship is the most important asset. Every OTA guest in your database is a potential direct booking waiting to happen — and the OTA winback campaign is the path to getting there.

Key takeaways

Your database grew 24% last year. But the more important question is whether the data inside it is complete, clean, and connected enough to power the decisions you’re asking it to support. In an AI world, the answer to that question carries higher stakes than it did twelve months ago.

Pick one gap from this post — phone capture, OTA conversion, PMS integration — and close it before your next campaign cycle. Database health isn’t a destination. It’s a practice. And the hotels that started earlier will have a decisive advantage when AI amplifies the decisions that run on it.

Ready to go deeper? Register for Webinar 2 of the Benchmark Leadership Series to hear directly from industry practitioners on what database health looks like in practice and what’s driving the biggest gains. Or download the full 2026 Database Health Benchmark Report to dig into the data yourself.

Frequently asked questions