Chief Commercial Officer Tan Bee Leng on the mindset shift that took the company from owner to operator – and the one she says is even bigger
Ask Tan Bee Leng, chief commercial officer of Ascott, what her immediate challenge is and her answer was not what I had expected.
“Parity,” she says, “managing that with our ‘frenemies’ – OTAs, wholesalers – and it’s so easy to compare prices these days. The competition is increasing and it’s a race to the bottom and extremely unhealthy for the industry.
“This is why I am pushing my team to look at uplifting experiences so that people won’t quibble over $10 or $20 more. And that differentiation happens when you book direct, which is why we are giving loyalty such a big push – we have to treat people who book with us directly, especially well.”
It is this single-minded focus on guest experience that’s also driving her into what she calls the “second and bigger transformation” taking place at Ascott.
The first transformation: from owner to operator
The first transformation was Ascott’s asset-light strategy – the Singapore-based, CapitaLand-owned hospitality company’s decades-long shift from a largely owner-operated model to a third-party management model at scale.
To pull it off, Tan and her team had to build an entire digital infrastructure to support the growth – a website, app, loyalty programme, distribution channels, the works, all needed to operate at scale.
“More than that, it was a mindset change – from owner-operated to third party management, from asset value to scale for third party, from asset appreciation to revenue and operation optimisations. I still remember how hard that was,” she laughs, ruefully.
It’s a strategy that has clearly paid off. Ascott signed a record 19,000 units across 102 properties in 2025, up 27 percent year-on-year, pushing its global footprint past 1,000 properties and 176,000 units in more than 230 cities and 40 countries. It operates 15 brands in total, of which around half have global footprint. Of its portfolio, it now owns only around one percent –essentially the mirror image of the starting point of its asset-light strategy.
The second transformation: orchestrating Agentic AI
Now comes the second transformation – orchestrating Agentic AI across its platform to pull together a guest experience that bridges pre-stay, in-stay and post-stay, in particular for its Discover ASR loyalty members, reaching 10 million since its launch five years ago, and guests who book direct.
The rationale is simple: book direct, be loyal, and Ascott will give you a superior guest experience throughout the whole customer journey.
“We are good at the pre-stay, we can do better in-stay and we can definitely do better with post-stay – this is where most hotels fail and yet it’s the last memory guests have of us,” says Tan. “With Agentic AI, we have the opportunity to pull the pieces together and orchestrate it into a personalised guest experience.”
Of course, she knows, it’s easier said than done.
“The challenge is all the parts are sitting in different silos, they were designed and written for humans, and now we have to rewrite them for machines and put them into a common language which hopefully will still be relevant as we build,” says Tan. “The technology is changing so fast, there may be a new language out by the time we build and so we have to be able to pivot easily if that’s the case.”
I likened it to the evolution of video technology – VHS to Betamax to Blu-Ray, before a new streaming language arrived altogether.
Tan nods and laughs. “Yes, how do we keep up with the technology? Do we need to? In many ways, it’s a leap of faith. Blu-Ray was the better technology, but users didn’t switch and it became irrelevant.
“The assumption we are making is that travellers want to deal with machines. Right now, the evidence is they do in the discovery and planning stage – we are definitely seeing travellers use machines more in this phase – and even though they are not booking yet, the assumption is they will.
“And we have to be ready.”
And then there’s the matter of costs – how much do you invest in this leap of faith, particularly at a time of challenging market conditions when owners still expect their returns? How do you convince the CFO to commit resources to a “leap of faith”?
“I tend to break things down into bite-sized pieces – build it piece by piece, test, then add on before we deploy. It’s easier to make the case then,” she says, smiling.
Reworking the foundation layer: from SEO to GEO
The first piece is Ascott’s own content. “We are starting with how we describe all our properties and write it in a way that appeals to both humans and machines,” says Tan.
The second piece is external content – user reviews and social media.
“These two pieces will form the foundation layer around which machines will recommend us. We are monitoring how often we get recommended by the machines and we’ve seen a 450 percent increase in recommendations of Ascott by LLMs, albeit from a low base. But it’s working.”
The next piece is booking. Ascott is rebuilding its core tech architecture to support agentic commerce, including evolving its existing AI concierge Cubby – which has already handled 900,000+ guest enquiries since 2023 – from a conversational tool into a full booking agent.
Then there’s the loyalty piece and the call centre.
“All of this needs orchestration so that we can look after our loyalty members well – recognition means a lot to our guests. That is our end goal.”
An Accenture study that backs up the assumption
The partner Tan has chosen to work with on this is Accenture, which recently released findings from its Consumer Pulse 2026 survey showing:
- More than 1 in 4 consumers now use large language models as their number one discovery channel
- 9 in 10 would like to shop directly inside gen AI tools
- 8 in 10 consumers are likely to let gen AI tools influence 50 percent of their spend decisions
- 68 percent of consumers want agents to shop for their idealised self – supporting goals like making healthier choices, staying on budget or upgrading more intentionally. The focus is not optimisation, it is aspiration
- 57 percent would instruct their AI agent which brands to consider, showing that brand preference remains intentional and influential – but 1 in 3 would switch preferred brands if an AI agent said it was a better fit
- 85 percent trust AI agents over their best friends to make a purchase
Asked why she picked a consultancy over a tech provider for the project, she says, “I wanted the governance part as well as the human part. I was afraid a tech provider would just focus on the tech and would not give us the balance we seek. We are extremely protective of the human piece and guests ultimately pay for the hospitality experience.”
Accenture isn’t the only partner helping Ascott become Agentic-AI-ready. It is working with Amadeus to implement a new central reservations system built for attribute-based, AI-readable inventory, as well as EHL Hospitality Business School, which is training Ascott’s workforce to work alongside AI tools.
On hindsight, the first transformation – from asset-heavy to asset-light – has clearly worked, given Ascott’s growing footprint across Asia Pacific and beyond.
I asked Tan how confident she is about this second transformation, and she’s the first to admit there’s no certainty, because “everything is so uncertain and no one really knows how it will play out, but we have to move with our travellers and be ready.”
Asked what lessons she’s carrying forward, she says: “The human part is extremely important even if you involve machines. The success of the first transformation rested on humans working as connectors across the different programmes. This time, it will be humans powered by machines that will help us communicate with the systems.”
Tan Bee Leng will be speaking at WiT Singapore, September 30-October 2.
Get your tickets now.
