41.7% of global short-term rental bookings are now made within just seven days before arrival. It was one of the key findings presented at the Short Stay Summit 2026, held in London on April 22.
The event brought together more than 1,300 industry professionals and major players in the global market, with a focus on Airbnb, Booking.com and Vrbo.
Last-minute bookings: The new normal
The booking data, presented at the conference by Smoobu, confirms what many property managers are already experiencing: the market is becoming increasingly unpredictable, with shorter booking windows requiring far greater operational flexibility.
For hosts, this translates into continuous pressure to maintain last-minute availability, manage rapid turnovers and check-ins, and rely more heavily on dynamic pricing and real-time data.
In other words, operating a short-term rental increasingly resembles running a real-time business rather than a predictable booking model.
A two-speed industry
Hosted at London’s historic Old Billingsgate, the Short Stay Summit focused heavily on the future of the sector: artificial intelligence, revenue management tools, direct bookings and growth strategies.
Yet behind the polished presentations, a clear divide emerged.
On one side, platforms and technology providers are focused on AI, automation and new products. On the other, hosts continue to seek answers to basic operational concerns: payment delays, fee transparency and unreliable support.
That contrast was evident during the “Inside the Platforms” panel, where executives from major OTAs presented product roadmaps but offered limited clarity on everyday issues directly affecting host revenues.
Technology: Solution or distraction?
A significant part of the summit was dedicated to new technologies:
- Hostaway showcased solutions for direct bookings
- PriceLabs expanded into broader revenue management tools
- Key Data introduced AI-powered data products
While important, many of these solutions are geared toward larger-scale operators. Small hosts with one or two properties often remain outside the core design framework.
The reality for hosts
At the same time, the daily reality for hosts remains far removed from the narrative of industry reinvention.
Unilateral commission increases, terms allowing payment delays and controversial review systems continue to create uncertainty—especially for hosts relying on a single property for income.
This is where the 41.7% figure becomes even more significant: when bookings increasingly arrive at the last minute, any payment disruption or operational failure has an immediate financial impact.
A critical question for the sector
The most important question raised at the summit was not about technology or growth. It was more fundamental:
Can the sector continue evolving around platforms and data when its operations still depend on thousands of small hosts struggling with day-to-day realities?
The Short Stay Summit provided many answers about where the market is heading. It offered far fewer answers about how those sustaining the system can remain viable.
And as booking windows continue to shrink—as evidenced by the 41.7% figure—that gap may widen further.
AI in short-term rentals: Big benefits, rising costs
Artificial intelligence was another major theme at the summit. The conclusion was clear: AI offers significant efficiency gains, but the cost of using these services is expected to rise sharply. If that happens, many hosts may struggle to afford them.
Steve Schwab, founder and CEO of Casago, and Graham Donoghue, CEO of Forge Holiday Group, both argued that the cost of so-called “tokens”—the units used to access AI models—is likely to increase significantly as investors seek returns on billions already poured into the sector.
Donoghue revealed that his company already spends between $30,000 and $50,000 per month on tokens through Anthropic and is even considering adding a finance role focused solely on optimizing that spend.
Schwab described the emergence of a “token economy,” noting that tokens currently represent only 20% to 30% of total computing costs.
“We are moving toward a reality where compute cycles could become extremely expensive and investors will demand profitability,” he warned, suggesting that without major changes in AI architecture, access to the technology may remain difficult at scale.
Productivity gains are real
Despite rising costs, companies are already seeing tangible benefits.
Donoghue said Forge began adopting AI about a year ago, initially to automate repetitive tasks such as image processing and call transcription.
Today, use cases have expanded into more advanced applications. One example is “Vivid Stay,” a tool that converts static property images into video and drone-style visual content.
“A process that used to cost up to $3,000 can now be completed in 25 minutes,” he said.
Tasks that once took weeks or months can now be completed in days, significantly improving productivity.
A warning for smaller operators
One notable piece of advice for smaller businesses was to avoid excessive experimentation with “vibe coding” without clear access to APIs and data.
According to Donoghue, many such projects reach 80% completion but never get deployed, resulting in wasted resources.
A threat to OTAs?
At a strategic level, both executives suggested AI could eventually drive disintermediation for major online travel agencies.
Donoghue argued that large AI models may offer more personalized and relevant travel recommendations, potentially reshaping how consumers search and book trips.
Schwab went further, suggesting technologies are already emerging that could challenge the dominance of traditional OTAs and transform the search landscape altogether.
The big picture
The broader conclusion is that artificial intelligence is rapidly transforming short-term rentals and hospitality, creating significant opportunities for efficiency and innovation.
But rising costs, organizational complexity and technological constraints are creating a far more challenging environment—one where operators must balance investment with long-term sustainability.
