Back in 2004, when finding a decent travel deal online was a time consuming task involving several booking and airline sites, Paul English and Steve Hafner founded KAYAK, a travel search engine designed to find the cheapest deals on flights, hotels, and car rentals. 

After 10 years of turning the company into the final word in travel booking (seven years before Google Flights was launched), KAYAK was sold to Priceline for nearly $2 billion. For English, a serial entrepreneur by nature, it was building the product and recruiting those vital first employees that excited him most.

English has founded and sold five companies to date — his latest, travel tech startup, was just acquired by Capital One. Aside from assisting with the transition as an entrepreneur-in-residence, English has shifted his focus from travel to podcasting. 

His newest startup, Moonbeam, which he co-founded with CEO Mike Chambers in June, is a podcast discovery app that functions a lot like TikTok. An avid podcast listener himself, English became fascinated with TikTok’s eerily accurate algorithm and began toying with the idea of combining the two. If TikTok could figure out English’s love for dogs and bass players, why couldn’t Moonbeam do the same for podcasts?

We spoke with English just before Lola’s acquisition announcement, and discussed what Moonbeam learned from TikTok’s algorithm, why building a good team should be every founder’s first priority, and the best podcasts he’s discovered on his app.

This interview has been edited for length and clarity.

The Business of Business: So first off, I would love to learn more about Lola and what got you interested in joining the company.

Paul English: We started in 2015, so six years ago now. And the initial idea for Lola actually was to build a mobile app that allows you to work with a remote assistant efficiently. So give them access to your calendar, to your contacts, to your credit cards, to email if you want that, and have a simple way of organizing communication and access to your information. 

Very early on one of our investors said, “This sounds interesting, but why don't you do it for us for travel?” And I thought, I was never going to do travel again, because I had spent 10 years at KAYAK. I started that in 2004. So I wasn't intending on yet another travel company. But it sounded like an intriguing idea. So we began that journey. 

Very quickly, we found that of all of our users at Lola in the first year, like 80% of them were using us for business travel, and we weren't aiming for business travel. But we found that business travelers are the ones who really need a remote assistant, a kind of remote concierge, and now began the journey of building business travel. And then ultimately, when the pandemic hit a year and a half ago, we did a radical pivot. Instead of going to CFOs and saying, buying local software will reduce the cost of your business travel, we went to CFOs and said, buying Lola software will reduce the cost of all your employees. And so we really did a hard pivot to become sort of a FinTech company selling travel and expense management. 

“When I think about my company and what I do every day, it's always team first, customer second, profit third.”

What does the future hold for Lola?

Well, the exciting news is that we're actually announcing tomorrow [Oct. 8] that Lola has just been sold. And as of tomorrow, we will become part of Capital One. Capital One has acquired us because they really wanted the FinTech technology, and they wanted the team. They really have an interest in building up a FinTech center in Boston, and they want us to help lead that effort. So we're going to be releasing a lot of the products that people used at Lola, particularly the last year and a half of the FinTech products will have new versions of that around business payments, and a number of other new products will be created for Capital One.

How much was that sale for?

We're not announcing the details of the sale. It was a very good outcome. The team is very, very happy.

Congratulations on that. I'm also curious as to what you Your role within the company is going to be going forward.

So I'm doing two things. I'm co-leading the Boston office with Mike Volpe. Mike has been the CEO for the last few years. And then I'm also acting as an entrepreneur-in-residence, which I think of their 50,000 or 60,000 employees. I might be the only ER at the company, I'm not 100% sure of that, but I'm going to be helping them look at new ideas, look at other companies they might want to invest in or acquire. And really just, I would say, ideation across FinTech, across all of the businesses, but in the short term working with Mike Volpe, I'm going to be very focused on delivering the new products for the Lola team, which we're proud of Capital One Boston for growing that organization.

You have other ventures that you're in the midst of working on. One of the more interesting ones is Moonbeam. What got you interested in podcasts in the first place?

So I've become an avid podcast listener. I've been listening to podcasts for I don't know how long, however long they've been out since app for a set of podcast app, 10 years, whatever it's been. But podcasts, as you know, have exploded over the last couple of years. My title generally for most companies is CTO. I go back and forth between CEO and Chief Technical Officer, as a CTO. The three things you're in charge of is product strategy, design, and then engineering, so the building of the software. If you look at what I actually did at KAYAK, and at Lola, I spend most of my time on product strategy and design and I'm very opinionated about products. 

So the products I use a lot when I get frustrated with them, as an exercise, I'll just redesign them just to think about how I would want Instagram to work or TikTok or whatever. And for the podcast, I was so frustrated with the Apple app in particular, I said, We need a better app, I want an app with a better user experience. And then once we have that app, there's two things we're trying to solve. One is we want to make it easier to discover new content, because how can you possibly find interesting content when there's millions and millions of podcasts out there? We want to build a system that’s based on human curators in machine learning, to help you find better content based on your requirements. And then the second piece of functionality we're building is we're really interested in the relationship between the podcast host and his or her listeners, and so you're building a number of tools to let them interact.

What kind of tools are those?

So for the second group, between the host and the listeners, right now, Moonbeam allows you to create a curated set of clips of all the best podcasts. We have two minute clips, sometimes shorter, sometimes longer than that. We allow hosts to create their own curated clips to share those with users. We allow users to tip posts with one click. If someone's inspired by something on your show, they can click a button and tip you $1 or $10, whatever they want. We're adding the ability for hosts to email listeners of their podcast, that's an opt-in for the user if they want to hear if they're excited about your show. And they want more from you, like I need seven days a week, I can't wait for a show through Moonbeam. There will be a mechanism where you can actually email them once a week or as often as you want. And then the last thing we're adding is discussion forums so that you can see how users respond to particular moments within your episodes.

You’ve said before that Moonbeam took a page from the book of TikTok. So those two minute clips sound a lot like that. In what ways are they similar

So like a lot of people, I became obsessed with TikTok. It's interesting. The thing I use most on my Apple TV is YouTube. And the YouTube recommendation algorithm is just terrible. So for example, a few months ago one night I was on a Prince kick — I really like Prince as a musician. Watched a bunch of his videos. The guy's a genius. He's amazing. It’s now three months later. I haven't watched a Prince video in probably three months. Every time I open YouTube on my Apple TV all I see is Prince videos. Like how dumb are they at Google that they don't know how to streamline YouTube? It's terrible. 

However, if you look at TikTok, when you first use TikTok you get people dancing, music videos, that's great. But the more you use it, the more it tunes into other things, figures out what you're interested in, and for me if I have half an hour to kill and want some entertainment, TikTok is by far the best way for me to deliver content I'm interested in. So TikTok discovered I like dogs. It's discovered I like bass guitarists. I follow a lot of bass guitarists now and I'm a bass guitarist. I'm a pretty bad bass guitarist, but I like watching really professional, great guitarists. I like watching up and coming musicians. It's been fun to watch people who are not very popular TikTok new accounts and then blow up and suddenly have a million or 10 million or more listeners followers for that. And that's really fun to watch. 

“If you look at what I actually did at KAYAK, and at Lola, I spend most of my time on product strategy and design and I'm very opinionated about products.”

And with Moonbeam, we're trying to do a similar thing to what TikTok has done, which is we have a user interface which is instrumented to show intent. And we watch things such as: Do you listen to a clip all the way through? Do you share it? Do you comment or read comments? Do you double tap it to like it, to favorite it? Do you subscribe to that episode, do you tip the podcast hosts? We then look at the text of the transcription of the episode to pick out keywords of what you're interested in. So if you listen to a lot of clips that talk about Elon Musk, we probably say, okay, maybe this person is interested. And last, we'll try another one and another one. So the idea is, the more you use Moonbeam, the better that recommendation system is going to become. And we're still in the early stages, but it doesn't work perfectly yet. But we're very focused on it. And we really want to be the best podcast discovery app out there.

Speaking of algorithms, and giving people what they want maybe before they realize they want it, TikTok’s algorithm in particular gets described as uncanny. To your knowledge, what makes that algorithm different from YouTube’s? How do you make an algorithm that uncannily good?

I mean, how bad YouTube is, is kind of shocking to me, because I do look at Google as one of the best AI companies in the world. If thousands of engineers work on machine learning and AI, I just don't think any of them work at YouTube. I just don't get it, like how bad it is. But the other thing is TikTok has more signals than YouTube. So TikTok, you're watching a video, you swipe up to go to the next one, you know, you swipe up to keep flipping through, and you interact and you touch. It's like this tactile UI, double tap to like it, click one button to share it, you can read all the comments, you can add comments. And it's really watching that interaction, that they're taking all of those signals and feeding it into the algorithm. And quite effectively.

Right, every little movement, every little behavior adds up to that, even clicking on the comments. KAYAK was designed to give better flight deals than anyone else on the internet at the time. It's almost as if Google Flights is the main competitor known for getting the best deals. What got you interested in these algorithms for finding travel deals, and what got you interested in travel software to begin with?

So I've started five companies now and actually, as of this morning, have sold five companies. So I’m five or five right now, which I'm pretty excited about. When I started KAYAK with Steve Hafner as my co-founder, the CEO — KAYAK was actually his idea, not mine. He pitched me, we had a chance meeting in Cambridge at a VC office. Steve and I met for drinks and he gave me the pitch. And I liked it a lot and gave him some feedback. I think that night I wrote up my thoughts about how to build a travel product. 

But even though the original idea was his, which is not a booking engine, but just a pure search engine, we found the flight you want, we'll show you all the places you can buy it and click the link, we’ll send you directly to,, whatever. That was his original idea. But I owned it immediately because I was so excited about it. And I thought this is a great idea because I'm someone who loves travel. I do about 100,000 miles a year. And before KAYAK, I had to go to four or five sites before I would pull the trigger on a deal because I had to search everywhere, because no one site had everything. And with KAYAK, the goal was we have everything. And the reason it was easy to do it is we didn't have to wait for commercial agreements for every airline or any hotel because doing that would take you more than a year. All we had to do was scrape the website and then send traffic to that website when people decided to buy that flight or rent a car. And I just had a total blast. It was 10 years of my life, a really fun 10 years.

Yeah, it sounds like you got roped into travel twice.

That's right.

What lessons do you still carry with you from that decade with KAYAK?

With each of my companies, when I think about my company and what I do every day, it's always team first, customer second, profit third. And that sounds like a terrible prioritization for my investors. But the reason I’ve been reasonably able to raise money with each company, I mean one is my companies are successful, but also I've kind of convinced investors that if you're the world's greatest recruiter in tech, and you can build that magical team that has the mojo, works hard, loves working, loves a problem has space to work on it, does really rapid innovation, magical teams create magical products. And with each company, I study the art of recruiting and interviewing and hiring. And with each company, I tried to get just a little bit better at it. I learned from people at my company. So I have a core group of people that have followed me and kind of come to the company, some people I've worked with for 30 years now. 

“How bad YouTube’s [algorithm] is, is kind of shocking to me, because I look at Google as one of the best AI companies in the world.”

But I also am very intentional about when I start a new company, I don't want to, let's say, take the KAYAK team and have the KAYAK team work on the new idea. I want fresh blood and fresh ideas. And it's been interesting for me, because when I hire someone that's working in a different industry, and that I haven't worked in before, I learn something from them. And a lot of our learnings, particularly how it relates to recruiting, and recruiting good leaders, the beginning, a lot of learnings about, okay, how to work together to build the next 10 people and 50 people and 100 people. And I just love that problem.

What’s your advice for entrepreneurs out there, maybe for someone who might want to be a serial entrepreneur and sell five companies?

Yeah, the first thing I would say is, however many hours you think you're gonna work a week recruiting, double that. And it sounds like how do I have time to recruit when I'm busy writing my business plan, meeting with venture capitalists, designing a product, writing code? But I'm telling you, if you're really real class recruiting, and hire magical people around you, you'll get the right results. You'll build products that blow away customers, and you build a business that has extraordinary revenue and profit.

Since Moonbeam is your main focus at the moment, what is your go-to podcast right now?

Oh, that's interesting. As I mentioned earlier, I was just on Guy Raz, How I Built This. That was a real delight for me because I’ve been listening to his show for a long, long time. When I think about more obscure shows that I'm addicted to now that I found through Moonbeam. Moonbeam just randomly starts playing these two minute clips. And it narrows down on finding things you like. I'll name a few podcasts that Moonbeam has discovered for me. 

One of them is a history of rock and roll in 500 songs or something like that. It's amazing, every episode. It's called A History of Rock Music in 500 Songs. And it's a very cool show, each episode just does one song and tells you kind of a backstory behind it. There's a show that I like a lot called The Writer’s Almanac with Garrison Keillor which talks about this day in history, and different authors. That's really interesting. 

There's another show hosted by an Irish poet, the show is called Poetry Unbound. And the cool thing about that is, read a poem, and then he'll discuss it and analyze it and then he'll read it again. And it's really cool. I like poetry a lot and it's really cool to have someone who's probably a lot more knowledgeable about literature and poetry than I am and interpret it. And then the second time he reads it, you just get new insights you didn't get from reading a poem once by yourself the first time. So those are three that I found through Moonbeam and I really enjoyed all three of those.

Ad placeholder