a submission by Martin Bookallil, CIO & CTO Advisor
The organisations most likely to disrupt your industry are probably not studying it.
They are watching an adjacent industry solve the same structural problem faster, then moving across.
Hospitality taught me this firsthand.
Hospitality is a business of fragmented data, high customer acquisition costs, complex distribution and operational systems that were built decades ago that never quite caught up with each other.
For most of that time, those problems felt unique to hotels.
They are not.
The same description increasingly applies to healthcare, logistics, financial services, retail and many service-heavy industries operating legacy enterprise environments today.
I spent 26 years leading technology across Asia Pacific for Marriott International, growing the region from 25 properties to over 1,200 across 27 countries. Since then, I have advised organisations on digital transformation, emerging technology and technology strategy across multiple sectors.
What hospitality taught me, sometimes painfully, is that the problems organisations believe are industry-specific are usually structural.
Those structural problems are becoming increasingly difficult to ignore.
The stack was built for a different era
For decades, enterprise technology followed a familiar pattern: operational pillars, each doing its job, connected through APIs and integrations that were rarely as seamless as vendors promised.
In hospitality, those pillars are PMS, CRS, RMS, CRM and Channel Managers.
In healthcare they are EMR, billing, scheduling and patient engagement systems.
In logistics they are ERP, WMS, TMS and carrier platforms.
The names differ.
The structural problem is remarkably similar.
AI is beginning to expose those structures in uncomfortable ways.
Some vendors are layering AI capabilities onto existing architectures, while others are reconsidering the architecture itself.
The industry is still determining whether AI is simply another enhancement layer, or whether it requires a deeper architectural rethink.
The question technology leaders should increasingly ask is no longer:
“Which platform has the best AI roadmap?”
It is:
“Does AI require a fundamentally different operating model, and if so, how do we get there from where we currently sit?”
Large organisations face a genuine constraint here.
Major transformation programmes consume years and significant capital before the technology landscape shifts again.
Those investments create legitimate caution.
Change will happen, but for large organisations it will happen gradually, and that is not entirely irrational.
The customer has already moved
The most urgent disruption is not coming from internal operations.
It is coming from the front door.
In hospitality, guests who once searched Google, browsed an OTA or visited a brand website are increasingly opening AI tools and asking:
“I need a hotel in Singapore with a rooftop pool near good restaurants suitable for a family.”
But this pattern extends well beyond hotels.
Patients ask AI which specialist to see.
Procurement teams ask AI to recommend suppliers.
Consumers ask AI to compare products and services.
Traditional discovery pathways through search engines, marketplaces and direct channels are beginning to compress or disappear entirely before the customer reaches the organisation at all.
I have observed agentic AI platforms in travel verticals making pricing and distribution decisions in milliseconds, drawing simultaneously on weather patterns, macroeconomic signals and demand indicators.
Platforms of this type are reporting meaningful revenue improvements following deployment.
Those approaches are now moving across industries, and the organisations behind them are not waiting for an invitation.
The most useful question right now is not:
“How do we improve our search rankings?”
It is:
“How visible and accessible is our organisation in an AI-mediated environment, and how does our customer acquisition model adapt when the decision may already be forming before the customer reaches us?”
From systems to intelligent data layers
Across conversations with CIOs and technology leaders, I increasingly hear the discussion shifting from “which platform should we buy?” to something more fundamental:
“How do we create an environment where operational, customer, market and external data can work together intelligently?”
This is a data pool question, not a software selection question.
Across industries, information often exists somewhere in the ecosystem but cannot be surfaced when it is needed.
In hospitality I saw this repeatedly. A guest leaves detailed feedback on an external platform. The information exists, but the operational systems never see it, and pricing, service or loyalty decisions made the next day occur without it.
Multiply that across thousands of data points and dozens of systems and the true cost of architectural fragmentation begins to come into focus.
Earlier in my career, even a localised feature request could take six months to move through vendor development cycles. Today, similar changes can be prototyped in hours.
That difference in pace is not a minor efficiency gain.
It represents a different way of operating.
The organisations making the most progress are not simply buying new platforms. They are rethinking how information should flow before deciding what technology should enable it.
The economics of custom development have shifted. More than most realise.
This is where I think the scale of change is still being underestimated.
Recently, working largely independently, I built two workflow platforms tailored to my own business processes using AI development tools.
Total cost: approximately USD 1,500 in token spend.
I have led projects and reviewed vendor proposals where equivalent implementations of USD 700K up with ongoing maintenance commitments.
That was not considered unusual.
In enterprise technology procurement it was often considered reasonable.
The gap between USD 1,500 and USD 1M is not a marginal efficiency improvement.
It reflects a structural shift in software economics.
For most of the last two decades, technology scale advantages strongly favoured large enterprises.
The capital required for sophisticated custom systems was simply out of reach for smaller organisations.
AI may begin reversing that equation.
Smaller organisations can increasingly access capabilities that historically required large budgets, large teams and long development cycles.
That shift in competitive dynamics may ultimately prove more important than any individual AI application.
I am not suggesting organisations should suddenly build everything internally.
Security, accountability, integration complexity and operational continuity all matter.
But the default instinct to buy off-the-shelf because bespoke is prohibitively expensive deserves serious re-examination.
That calculus has changed.
Where to begin
Audit where critical business data lives. Most organisations discover fragmentation faster than expected, and that audit is the necessary first step before any technology decision.
Run AI-native experiments outside core systems. The operational risk of piloting a workflow tool in a non-critical area is low. The cost of not building organisational understanding of these tools is high. Start somewhere that can fail safely.
Reverse the sequence of your technology projects. Start with the business process, identify the friction, map how information should flow, then decide whether to buy, customise or build.
Prioritise pilots over programmes. A well-designed experiment that generates real organisational learning is worth more than a large capital commitment made on today’s information.
The organisations making the most meaningful headway right now are often not moving fastest. They are learning fastest.
The question worth sitting with
If you designed your technology environment today, with what you now know and what is now possible, would you build what you currently run?
For most large organisations, the honest answer is no. The interesting question is what they choose to do with that answer.
Large organisations are slow-moving ships for understandable reasons.
Smaller organisations without the weight of major legacy commitments will move first and demonstrate what is possible.
When they do, the pressure on larger organisations to reconsider will grow, not from vendors, but from competitive reality.
The organisations best positioned for the next decade will not necessarily be those with the largest technology budgets.
They will be those willing to look at their own industry with the same fresh eyes that outside entrants are already bringing to it and begin building toward a different architecture now.
That is where the conversation begins.
It rarely begins with technology.
