Research Briefings

Keep your PMS

PMS adoption does not remove the need for a stronger owned hotel layer. It changes the problem: keep the systems that transact the stay, then protect the room meaning, guest context, conditions, and human confirmations that can disappear between them.

Dan MetilleFounder of TamagaPublished by Tamaga · August 14, 2026

Research Briefing: why hotel-tech saturation changes the integration problem

Independent hotels do not enter 2026 with empty dashboards.

Many enter it with several systems already doing important work.

A 2025 study of more than 1,500 hotels in six European countries found that 75% used a property management system and 63% updated rates and availability through a channel manager. The same study counted more than 70 PMS brands in use: high adoption, but a fragmented technology landscape.¹

HOTREC’s 2024 distribution study, based on observations from more than 3,000 European hotels, found that most hotels already used internet booking engines, while smaller hotels remained particularly dependent on online travel agencies.² Cloudbeds’ 2026 dataset, drawn from 90 million bookings across tens of thousands of independent properties in 180 countries, reported a 63.4% OTA share of observed bookings in 2025.³

The market signal is not subtle:

The next hotel-technology problem is increasingly not how to install a first system. It is how to stop several useful systems from turning one stay into several different truths.

That changes what a new hospitality platform should try to own.

The hotel-stack paradox

The more mature a hotel’s transactional stack becomes, the less reason there is for another product to replace it.

A working PMS may already manage inventory and reservations. A booking engine may already transact the direct booking. A channel manager may already synchronize distribution. An RMS may influence rates and restrictions. A payment provider may already know whether money moved.

Replacing those authorities simply to improve the website can create more migration risk, more training, and another version of data that was already being managed somewhere else.

But a mature stack creates a different problem.

Consider one sentence:

Later arrival after 20:00 requires written confirmation.

The PMS can know that a room is reserved.

The booking engine can know which rate was selected.

The channel manager can know that inventory was reduced.

The payment provider can know that a deposit succeeded.

None of those facts, by themselves, proves that the condition requires written confirmation survived the journey, that the guest still understands it, or that a named person is responsible for resolving it before arrival.

Six systems can transact one stay and still lose one sentence.

That is the paradox.

Transactional saturation, semantic scarcity

Tamaga uses transactional saturation and semantic scarcity as analytical terms for this briefing. They are not established hotel-industry metrics.

Transactional saturation describes a property where the principal commercial and operational transactions already have credible systems of record.

Semantic scarcity describes the remaining shortage of governed meaning between those systems: what the room actually fits, which limitation matters, what the guest relied on, what remains conditional, and which human decision is still needed.

A simplified authority map looks like this:

Existing authority	What it can legitimately know	What it does not necessarily preserve
PMS / CRS	inventory, reservation and stay state	why the guest chose this room or which public condition they relied on
Booking engine	booking interaction, presented rate and booking result	the complete decision context that existed before handoff
Channel manager	rates, availability and distribution state across connected channels	whether a rewritten hotel condition kept the same consequential meaning
RMS	pricing recommendations, restrictions or revenue decisions according to the property’s stack	whether the guest-facing explanation still expresses that restriction honestly
Payment provider	authorization, capture, refund and settlement status	whether an operational promise attached to the stay has been fulfilled
Human host	judgment, exceptions and the ability to confirm what software cannot	a durable record unless the system gives that judgment somewhere to live

This is why “source of truth” is too blunt a phrase for a modern hotel.

A connected hotel needs an authority map: one legitimate owner for each kind of truth, plus an explicit path for meaning that must cross the boundary.

The missing layer is not another PMS

The evidence does not suggest that independent hotels need one more product to rebuild everything their stack already does. Tamaga’s own 2026 review of small-property technology found the harder problems recurring around fragmentation, interoperability, operational fit, training, cost, booking friction, and the transfer of context between systems.⁴

Academic hospitality research has been making the interoperability point for years. Buhalis and Leung described hospitality as a highly interconnected ecosystem and identified standardized communication, interconnectivity and interoperability as core requirements and persistent challenges for smart hospitality systems.⁵

Tamaga’s interpretation goes one step further:

Interoperability should not mean copying everything everywhere. It should mean preserving the right authority and the consequential meaning across the handoff.

That distinction matters.

If a PMS already owns availability, Tamaga should not maintain a parallel sellable inventory.

If a payment provider owns settlement, Tamaga should not invent payment truth.

If a booking engine can complete the reservation safely, Tamaga should not force the hotel into a second checkout merely to control the interface.

Tamaga’s useful territory is narrower:

govern the room and policy meaning the hotel is prepared to publish;
explain whether a room actually fits the party, independently of date availability;
preserve the consequential context on which the guest relied;
keep a request pending until the right authority confirms or declines it;
make the required human attention visible;
inspect whether later representations kept or changed the meaning.

The established booking authority still runs the reservation.

The promise should not disappear around it.

Minimum sufficient integration

The usual integration story rewards breadth: more logos, more connectors, more fields synchronized.

For an independent hotel, that can be the wrong objective.

Tamaga proposes a different test: what is the minimum sufficient integration required to protect this guest journey without creating another master?

That may be:

a read-only availability query;
a prefilled booking handoff;
an opaque context identifier carried into the booking flow;
a signed callback or webhook after reservation creation;
a narrow reservation reference used for reconciliation;
an explicit human handoff where the vendor interface or the promise itself does not justify automation.

A read-only connection is not an inferior integration when read-only is the correct authority boundary.

A deep link is not a failed API when the booking engine should remain in control.

A human decision is not missing automation when the hotel genuinely needs judgment.

Integration depth should follow the promise, not the sales page.

Why the owned layer still matters commercially

Keeping the transactional stack does not mean accepting that the owned website is merely decoration around it.

Cloudbeds’ 2026 report found OTA dependence increasing among the independent properties in its dataset.³ Separately, SiteMinder reported that hotel websites generated an average booking value of US$516 in 2025, versus US$312 through OTAs in its platform data; SiteMinder attributes the difference to higher-value rooms, longer stays and extras, and the figures should be read as vendor-platform evidence rather than a universal causal effect of booking direct.⁶

The two datasets point in the same strategic direction without proving the same thing: the direct channel remains commercially important while independent hotels operate inside increasingly complex distribution systems.

The owned layer therefore has a job bigger than presenting photographs before sending the visitor elsewhere.

It should help the traveler make a better decision before the transaction, carry the important context through the transaction, and give the hotel a way to act on what still needs judgment after the transaction.

That is a different job from inventory management.

The question to ask before buying another system

Before replacing a PMS or adding another large platform, take one guest promise and trace it.

For example: four proper sleeping places, connecting rooms, step-free access, a cot, airport transfer, parking, breakfast, dietary accommodation, or late arrival.

Then ask:

Where is the hotel’s established position recorded?
Which system is authoritative for the transactional facts around it?
What exactly does the guest see before booking?
What survives the handoff into the booking engine or PMS?
What remains a request rather than a confirmation?
Who has to act next?
Can the hotel later recover what the guest relied on?

If the existing stack answers all seven questions reliably, another layer may not be needed for that promise.

If it does not, replacing the PMS may still be the wrong repair.

The gap may sit between the systems rather than inside any one of them.

Tamaga’s position

Keep the stack that transacts the stay. Add only the layer the promise requires.

Tamaga Hospitality is designed around external authorities rather than a closed list of systems a hotel must adopt. The safest interface may be an API, webhook, callback, export, embedded flow, prefilled handoff, or read-only connection, depending on the property and the access its vendors actually provide.

Where deeper access is available and justified, the connection can deepen. Where a vendor limits access, the architecture should degrade honestly rather than simulate synchronization that does not exist.

The ambition is broad compatibility.

The discipline is narrower authority.

That is why the first question is not:

Which PMS should Tamaga replace?

It is:

Which part of the guest promise does the existing stack still fail to carry?

Evidence and scope

This Research Briefing combines external market evidence with Tamaga’s product interpretation.

The HES-SO figures describe a 2025 sample of more than 1,500 hotels in six European countries and should not be treated as a worldwide PMS-adoption census.¹ HOTREC’s distribution study covers more than 3,000 European hotels and focuses on distribution rather than the complete hotel technology stack.² Cloudbeds and SiteMinder publish very large current datasets, but both reflect bookings processed through their respective commercial platforms.³⁶

Hotel-stack paradox, transactional saturation, semantic scarcity, and minimum sufficient integration are Tamaga analytical constructs proposed in this briefing. They are not established industry-standard terms.

The current Tamaga Hotel Demo proves the governed House Record, Room Fit, Promise-to-Task, Host Attention, and Channel Lens mechanisms in a fictional, isolated environment. It should not be read as evidence that every vendor-specific PMS or booking-engine connector is already in production. Connector depth and status belong to the integration layer and must be stated explicitly for each implementation.

References

HES-SO Valais-Wallis. “European Hospitality — Revenue Management, KPIs and Distribution: A Fragmented Landscape?” 19 August 2025. https://www.hevs.ch/en/news/european-hospitality--revenue-management-kpis-and-distribution-a-fragmented-landscape--212307 ↩ ↩²
HOTREC. “Digital Trends in Accommodation: Hotels, Booking.com and DMA” and Hotel Distribution Study 2024. 2 July 2024. https://www.hotrec.eu/en/news_digital-trends-in-accommodation-hotels-booking-com-and-dma_C2_A0_C2_A0.html ↩ ↩²
Cloudbeds. The 2026 State of Independent Hotels. Based on 90 million bookings across tens of thousands of independent properties in 180 countries. Accessed 9 August 2026. https://www.cloudbeds.com/hospitality-industry-report/ ↩ ↩² ↩³
Tamaga. Small Independent and Family-Run Hotel Technology in 2026. Research working paper, August 2026. The report synthesizes vendor datasets, academic literature, official documentation and qualitative operator evidence, with explicit limitations on global small-property adoption estimates. ↩
Buhalis, Dimitrios, and Rosanna Leung. “Smart hospitality — Interconnectivity and interoperability towards an ecosystem.” International Journal of Hospitality Management 71 (2018): 41–50. https://doi.org/10.1016/j.ijhm.2017.11.011 ↩
SiteMinder. Hotel Booking Trends 2026. Hotel websites generated an average booking value of US$516 in 2025 versus US$312 through OTAs in the SiteMinder dataset. Accessed 9 August 2026. https://www.siteminder.com/hotel-booking-trends/ ↩ ↩²