Once Long Lake's $6.3 billion acquisition of American Express Global Business Travel is complete, which is expected in the second half of 2026, it may only take a couple of weeks for the business to begin to feel the margin lift that comes from its initial integration with Long Lake's Nexus AI platform, Long Lake co-founder and CEO Alex Taubman recently said on a podcast.
"Roughly 80 per cent of the infrastructure is shared across the verticals [we invest in] and then there's a lot of work to take it and deploy it into those end markets," Taubman told No Priors podcast host Elad Gil, adding that early acquisitions took "over a year" to find the potential of the AI and begin to see it in business outcomes. Now, however, he told Gil, there is almost "instant time savings" after which the company begins to look at growth over the longer term.
Should the Amex GBT deal go through, which is broadly expected, it will be Long Lake's 31st acquisition but its first to take a public company private – and it moved fast.
"If you look at the multiple, they paid almost 12 times the EBITDA," said mergers and acquisitions broker Matt Zito of Travelexits.com. "That's a really interesting piece that they paid that much above market price. But it's also really hard to do a transaction at that scale. ... They got everyone on board quickly. It was pretty impressive."
After a fast agreement, how fast do they think transformation will happen?
The short and the long of It
Long Lake has a penchant for what Taubman called "sleepy" service industries and applying its artificial intelligence platform to the companies' core value proposition. The expectation, as he put it, is that employees become more productive on AI-powered tools and provide incrementally faster and better service to customers, thereby attracting new customers and thus growing revenue. Taubman claimed that Long Lake, with this strategy, has become the fastest-growing company in homeowners association management – one of its best verticals – and has lifted annual growth in those companies from an original zero to 5 per cent to more than 20 per cent each year.
The plot may sound simple, but the narrative of "instant" margin lift has its limits.
Long Lake's Nexus platform deployment involves mapping workflows, understanding and cleaning up data sources and integrating them with the platform to make the data easier for the models to access. Taubman said that "takes a lot of customisation and significant applied AI engineering capabilities," and that's just the beginning of the engagement with the acquired company.
The longer tail of the work will involve engineers sitting day in and day out with Amex GBT, understanding the pain points and helping the company "fix all [its] problems" over the course of the next two years. But the real transformation actually takes longer, Taubman told Gil. "What we're doing is actually really hard. And … you can't do this in a year; you can't do this even in two years, three years. This is a multi-year transformation."
Public vs. private transformation
"In the macro picture, corporate travel is such a weird beast," said Thayer Investment Partners venture partner Cara Whitehill, whose previous roles have included executive positions at Expedia, Deem and Traxo. "It's one of those areas that is really hard for public sector investors to understand. … Unlike a basic booking on Expedia, [corporate travel's] different layers and stakeholders and components [make it] trickier to understand. And when you are dealing with this moment when AI can have so many applications, it makes sense to go private because the things you need to do as a company to build and take advantage of the technology aren't things that are going to manifest in the next quarterly earnings call."
Corporate travel technology consultant Steve Clagg of Clagghaus Consulting echoed Whitehill's sentiments.
"Amex GBT has been narrating AI for several cycles," he said. "Private capital could change their pressure dynamic that might have kept them underinvesting in what you could call 'the boring stuff' that actually may really be good investments. Areas like ticket accuracy, exchange handling, disruption mitigation, rebooking and all the mid-office automation – investing in changes there could be really good, but they don't sound like fireworks to public shareholders."
He added that Long Lake's acquisition and posture of a longer-term interest in GBT boded well for the TMC and could result in faster decision-making and increased "product velocity," which could engender significant confidence in the overall strategy. A robust re-platforming might also accelerate GBT's integration of CWT, which it acquired last summer.
"That has to be dealt with; if it isn't, they're going to be painfully dragged down. … They need a unified stack in order to get where they want to go," Clagg said.
An enhanced team, however, is unlikely to come without some organisational and product shakeups, with Long Lake clearly poised to bring in its own tech muscle to realise Taubman's promised "instant time savings" from the first re-platforming round.
A spokesperson for Amex GBT said that a change in ownership would "not have any negative impact on resourcing" and said it would reduce the costs of operating as a public company. They also said Long Lake's "dedicated team of engineers" would be additive to current resources. And while the company would not go into long-term goals, "the broad plan is to accelerate our existing strategy rather than change course, and there are no contradictions to our existing strategy, including Complete and our aggressive AI rollout." Complete is the combined travel and expense offering from Amex GBT and SAP Concur.
That said, GBT chief information and technology officer David Thompson announced he will exit the company on 31 May.
TMC commercial models and growth opportunities
Clagg predicted AI-driven assessments, reconfigurations and automation will force the TMC to grapple with the economic conflicts baked into commercial models, in which both the buyer side and the supplier side vie for their needs with the TMC. Long Lake will have to tease out these revenue streams internally in order to enable the AI models to run efficient and effective operations. In theory, Clagg mused, an AI-driven re-platforming could offer opportunities to rethink fee structures.
"They'll have to deal with the economic differences. Like, when they're paying for tokens and they're building solutions around outcomes that have highly variable amount of workload – how that agentic AI gets to the outcome. It's not a flat transaction model cost; it's very different. How Long Lake rationalises their cogs with the revenues … that's the thing that they have to figure out. It may be purely internal, but I would hope that they would bring some more transparency to it."
Indeed, he said, private ownership could help Amex GBT rethink fee structures specifically to focus more around outcomes than transactions.
Asked why she thought Long Lake took on such a large TMC for an overhaul over a smaller one that may have proven its thesis more easily, Whitehill did not hesitate: "Take Amex GBT with its giant huge book of business and optimise for that. Just a few tweaks in the right place can already get you such a monumental return just because they are so big."
But Whitehill also said Long Lake and Amex GBT will have to do more than tweak the algorithm and figure out how to parse out murky revenue streams.
Over the longer term she sees erosion of the TMC's value in an AI-driven future – not just for Amex GBT but for TMCs overall. That could be a big issue for Long Lake as it looks to grow the business and realise a return on the 60 per cent premium it paid for GBT over the stock price.
"A lot of the core things that TMCs have stood on for a long time in terms of being that booking infrastructure layer … it's no longer really relevant," said Whitehill. "Maybe for some of them – the percentage of trips that are multiple cities and complex with different suppliers – there is a place for service in that scenario. But even for disruption and re-booking, that's a feature, not a business model.
"Plus, you have to look at leakage – it's been an issue forever. I was brought on at Deem in 2010 to 'consumerise' the tools and solve the leakage issue. [The industry] has been trying to solve that all this time, and it hasn't budged. That should tell you something: Travellers are going to book where they want to book, and with AI it's even easier with all the LLMs. They will find what they are looking for without ever engaging with their corporate booking tool or agency, and companies can capture that data. That's last era's battle. So, if I'm a TMC, I don't know where that leaves me. All the AI sophistication in the world isn't going to solve that fundamental issue for GBT or anybody else."
Tackling today's challenges
That's a big-picture investor's viewpoint that most corporate travel buyers can't apply to their programmes at this moment. Especially if they are clients of Amex GBT, they may want to put initiatives in motion to buffer their travellers and programmes from potential disruption.
"You have to understand what the functional replacement or augmentation paths are for your programme," said Clagg. "You have to be really clear with milestones in the agreement, including outcomes for changes, which might be spend reduction, might be experience increase. And it all has to be accounted for. This is the time to start – before any changes take place."
Clagg added that large, influential buyers should begin to collaborate and influence upward now, if they can. "If you have a strong vision and set of requirements, run hard with them now." Barring that, prepare for transformation:
- Put a magnifying glass on your contracts. Look specifically at clauses like the change of control. Most agreements will survive change of control, but some may have notification or consent rights that are worth recognising and using as leverage to improve the client's position and cement commitments from GBT going forward.
- Continue or increase optionality. Understand the TMC market and the various innovation partnerships in play. Run serious evaluations of competitors to give you a negotiation posture and/or a bona fide back up TMC.
- Get baseline metrics in order. Lock in your metrics now before any changes come down the pike. Average speed to answer, ticketing accuracy, your exchange and refund cycle, rework rates. Know traveller NPS satisfaction rates and have them ready should any service or satisfaction issues arise.
- Re-confirm data governance and training rights. Where is data housed, how long is it retained? What's the data model; what's the disclosure of the model provider; how are models trained with client data, if at all.
- Drill into roadmap specifics and objectives. Understand agent augmented vs. agent replacement milestones and timelines; and define human escalation paths for AI-powered functions. Align these with your programme.
"This acquisition news is big; it's huge," said Clagg. "I mean, it's definitely not to be overlooked how big of a change this will be. It's either going to be a huge change or it'll be business as usual. If it's business as usual, though, Long Lake will not have succeeded."