Building for Every Edge Case, From Compilers to Check-In
In approaching the 13th edition of the Vacation Rental World Summit, we’re releasing a series of interviews to leaders, key players and highly motivated people in our industry whose companies make a deep impact in our space and will be present at VRWS.
The aim is to get you some insight into these leaders and the company they founded, manage or work for, as well as getting some behind the scenes on their thinking, their motivation and their drive.
Thomas Tewolde
- You manage a billion dollar asset class luxury building with 700 doors, you sit on the board of Cauridor, one of the largest Fintech companies building payment infrastructure for unbanked populations across West Africa, while running a Guest Experience & Payments platform for the short term rental industry. What does sitting across such different ends of the payments spectrum actually teach you about where friction really lives in a transaction?
When it comes to digital guest experience for the short term rental market, one of the most vital yet overlooked friction is the lack of providing a smooth payment experience. Offering guests the ability to pay through their preferred payment methods, drastically increases conversion and experience.
- You started as a software engineer at IBM, building compiler features for the Cell Broadband Engine. From working at that level, how did that shape how you built ChargeAutomation?
Compilers don’t forgive edge cases, miss one, and the program breaks. That mindset from IBM is exactly how I approached building ChargeAutomation: mapping every scenario a host or hotelier could hit across the guest journey. Check-in, data collection, compliance, ID verification, payments, deposits, upselling, communication, and automating the entire workflow, while still giving operators the right level of control over when automation hands off to a human.
- ChargeAutomation’s own data shows ancillary revenue growth as high as 490% year over year through AI-timed upselling. At what point does an AI system that is very good at knowing when to offer a guest something start to feel less like good service and more like being sold to, and how do you design against that line?
The line gets crossed the moment the offer stops being about the guest’s stay and starts being about our revenue. AI that’s “good at timing” isn’t the goal, instead AI that’s good at relevance is.
If a guest gets an early check-in offer because they’re arriving on a red-eye, that’s service. If they get the same offer because it’s 9am and the algorithm knows conversion rates peak then, that’s being sold to, even if the guest can’t articulate why it feels off.
The 490% ancillary growth is a byproduct of getting relevance right, not of pushing harder. If we ever had to choose between the two, we’d choose relevance every time.
4 You’ve operated short-term rentals yourself for over a decade while also building the technology platform. When your team proposes a new feature, how much does having actually owned and run the asset change what you choose to prioritize?
Owning the asset means you’ve felt the cost of every gap firsthand, a chargeback you couldn’t fight because you didn’t have the right verification on file, a guest who trashed the place because the screening was too soft, a booking that fell through because check-in friction pushed them to cancel. You don’t need a case study to believe those problems are worth solving. You’ve lived them.
It also makes me ruthless about what we don’t build. Plenty of features are impressive in a demo and useless in practice, because they add a step for the operator without removing a real headache.
You’ll have the opportunity to meet Thomas directly at VRWS in Lyon
Not Yet Registered for VRWS?
