My first startup is worth $15Bn now. My latest one failed though. Here is why.
A story about building and breaking a travel startup.
I’ve been involved in many other startups since, as an investor or partner, but a year or so ago I went all in on a new one: Beachfix.
Before I started Beachfix, I had a jumble of thoughts in my head…
- First, it annoyed me that EVERY single major website in travel insists on asking you where you want to go, but very little to help you figure out the answer. Research involves roaming around the net browsing a ton of sites with 20 different tabs open. Surely there is a better way?
- Second, that there is no website on the planet which documents all the world’s beaches. Surely people would find such a site useful? When people go on holiday…they want a nice beach right?
- You might find a beach in Spain that is quiet, has good cuisine, clear water and calm water. You can also find such a beach in Portugal. It isn’t the country that determines whether the beach fits, but the specific destination and beach. But people often search by country.
So I thought, how about a website where you could come and tell me say ‘I want a beach holiday in a 4 star hotel for less than £1200 for 2 people, in March where it is at least 26 degrees and less than 8 hours flight away. I want clear water, a quiet beach and great food with also some good diving nearby.
Sounds great right? Yes. That’s what everyone told me.
I thought great, so people like the idea. But will they use it. And can I monetise them often enough, and for enough money?
The latter question particularly came up from conversations with people who had more insight into the space, like Sean Seton-Rodgers from PROfounders Capital, Alex Bainbridge, Peter Ward, Francesca Ecsery and Andrew Lacy — the latter four who had built startups in the space. It was clear from the conversations with them all that startups focusing on ‘discovery’ ie helping people research and select holidays had struggled to make money even when people engaged a lot with their site.
So I knew we were in a tough space. Ironically I had gone for travel because I wanted to be a bit contrarian. The startup scene was focusing on AI, VR and other ‘hot’ areas. I liked the idea of going against the grain. But there is a reason why consumer online travel isn’t hot….well at least not the bit I was in. Some other companies have still managed to find some great areas of opportunity — such as Much Better Adventures.
I also had thought there was one thing which might give us a chance — our focus on beaches. I’d seen other sites try this, but they did it for many types of holidays. We were focused on beaches only. We would be specialised! And we would have more and better data than anyone else! We would have a chance of helping people make decisions better than others had.
So we built a functional site early 2016. At this stage it was the multi-talented Maria Stylianou (tech + a whole lot more…and an ex-Entrepreneur First alumni. She is a techie who gets business) and myself. We also had interns who researched beaches for us. We collected over 100 data points per beach.
The metrics from the first site weren’t great. In fact we got one booking in 2 months. That’s it. To be fair, we didn’t spend a ton on marketing. But we spent enough to get a decent sense of the metrics. This version of the site can be found at http://old.beachfix.co/
What was encouraging though that 50% of our users via google ads — where we knew people were actually currently looking for a holiday — were filling in a 3 page form to get some results. That’s a very high completion rate!
So we knew people would use it. But monetisation was rubbish.
We did know we lacked a lot in that site. We didn’t have the data to filter by budget. So we would return results for holidays for £5000 next to holidays for £400 not knowing what users budgets were. This was a big user experience issue.
We also had zero hotel info on the site, relying on sending people out to booking.com etc to go look at hotels once they had decided they liked the destination. What wasn’t good was that not only were people not buying, but they weren’t even going to the booking sites to look at the hotels. This suggested our beach information pages weren’t good enough. Besides the lack of hotel data, it was also a bit cluttered and was information overload in how we had laid it out. We were so desperate to inform the users (and show them how much amazing info we had collected! Look how much we know about this beach!! Love us…love us!!) that we threw all our data and information at them in one go!
Around this time we bought in Nuno Veiga. Nuno came via a recommendation and I thought were just hiring a great marketer. After the first interview, I wanted to see how much he engaged with what we were doing and to come back with some thoughts about the product, brand etc. He came back to the second interview with 8 pages of thoughts! And good thoughts too!
After he joined though….it turned out he is also an expert and managing product development, and has great tech knowledge to boot! He had downplayed all this because he wanted to focus on marketing. But he was absolutely willing to get his hands dirty. And he did.
So…we rebuilt the site. We bought in a designer via Nuno — Jason Howmans — whose thinking as a designer was very interesting. He clearly thinks deeply about it. He is also a great front end developer. To help with the back-end we also bought in Luis Fernandes who is a serious serious techie. This guy lives and breathes tech.
We delivered a new site. And a beautiful site it is. Check it out at www.beachfix.co. At the same time I decided if this didn’t deliver improved metrics to suggest there was something happening here, I would wind the team down.
And again, the metrics disappointed despite the vastly improved user experience. Again people engaged with the product — some users spent hours on the site! — which is a lot for a site like ours.
Watching users on FullStory though was what made me realise the mindset of our users. We spoke with users, and we did lots of user testing. But the only way you really get what users think is watching users who don’t have any connection with you whatsoever use the site.
And the biggest emotion I felt from watching them was….indifference. They were just browsing.
It’s not that we hadn’t delivered a site they found reasonably interesting. They engaged with the site pretty well.
But we realised that the stage of the user journey we were in — discovery — users aren’t looking to make decisions. Or get to the answer quickly. Despite what loads of ‘users’ told us. They just wanted to browse.
The industry has statistics like how users spend 40 days picking holidays, and make 38 website visits before making a booking. We had dreams of cutting this down.
But I don’t think users (who will tell you otherwise) want a site to answer their holiday prayers in one session (except maybe last-minute hunters, but they are very deal-focused — not inspiration — when searching). They need to look around a lot. They want to understand what all the various options are before they book. The users lied to us! (not intentionally of course!).
The users lied to us!
And when they are ready to book, they have well-established habits to go make that booking. They go to booking.com (who owns 60% of the booking market for hotels in Europe — and only ever give you commission for bookings when a user clicks through to their site and books at the very same time….which is a problem. How many people book a hotel for a major holiday the first time they check the hotel out??) and Expedia. Or elsewhere.
So it turns out a lot of what I was told was right. It is difficult to monetise users with travel discovery. Paul Smith (who had some excellent thoughts he was kind enough to share with me a few months back) wrote this post. Highly recommended if you are in, or looking at the travel space.
We also now know why every major travel site asks you where you want to go, but not much to help you find the answer. Because the money is not in discovery, but in booking.
So, we spent about 15 months on Beachfix. We learnt a ton. We met a lot of great people, and we are extremely grateful to everyone who gave time to offer their insights on the space. It was my first B2C tech business as a founder, and it was a very instructive experience.
When I started out, I made no presumption in my head that we were guaranteed to succeed. I wanted to find out though if this could be made to work. I looked upon it as an experiment — and in that respect it was successful. We got a conclusion.
So, we have dissolved the team. Nuno was offered another job right away by another entrepreneur who had the chance to get to know him the past year. Maria will doubtless also be hoovered up with her talents. We also had a kick-ass intern Anna Ling. She will be a lawyer in less than a year, but in the meantime were grateful to have her — and her alpha personality — in the team.
Myself? I will take some time out, go skiing for a bit (I’m not very good — and I want to rectify that) and then start getting to know the AI space. I have already started doing the Machine Learning course on Coursera.
We will keep the website up though — and still work on some improvements on the side. So you can still find your perfect beach on there if you need some inspiration!
Although BeachFix didn’t make it, I view 2016 as a successful year in some ways. The personal growth, the connections — and despite its lack of monetisation — the product we built all leave me with a sense of satisfaction from the experience. Not 100% satisfaction of course. But something positive to take away as we all move on to the next chapter of our careers.
Whether it’s via feedback on the site, thoughts and ideas about improvements, strategy discussions or simple expressions of support — we had a ton of people help us out. So many people offered their insight and time — too many to mention. It’s humbling that people were so generous to us in this way. Thank you.
I’d like to finish by thanking the three co-working spaces that hosted us in our time with BeachFix. Runway East (cheers for the intro Daniel Murray!), Travel Tech Lab and Soho Labs (the latter closed now unfortunately). Runway East and the Travel Tech Lab are both fantastic and very supportive environments to be in as a startup. I highly recommend checking them out if you are looking for somewhere to base yourself.
Here’s to a successful 2017….