Madeline Parra, CEO and Co-Founder of Twizoo – interviewed January 2016
Twizoo is a restaurant recommendation app, it’s like a Yelp or a Trip Advisor, except all of our reviews are tweets.
I was previously working at Glaxo Smith Kline, working in social media monitoring analysis. While I was doing that I was noticing that in general reviews are basically broken. I mean, no one really leaves reviews any more, I don’t know anyone who goes online and leaves reviews on Trip Advisor, for example. I realised people are much more likely to share their experiences about a restaurant, a hotel or even a book on Twitter, Facebook or Instagram. So I wondered if there was a way to collate those mentions and bring them together. We built a filter that sits on top of Twitter and hears every mention of a restaurant or bar, either by the handle or just the name in the text. We then pull that Tweet in and analyse it to see if it is positive or negative and if the person is a customer or someone who works there, like a chef and we also look at any replies and if those replies agree or disagree to try and determine whether that’s a positive or a negative. That then builds a up a rating of every place based on what people had been Tweeting recently. Then the users of the app can search by area, for example ‘breakfast in Soho’, or ‘cocktails in Shoreditch’ and find out what’s been getting the most Tweets, good and bad, so what to avoid.
It’s actually a long story of how I came to this point, I was originally a teacher back in the States, where I’m from. I was teaching Math in inner city Baltimore. I don’t know if you’ve seen The Wire, but it was a lot like that! Pretty hard. I thought; this isn’t for me, although I didn’t really know what I wanted to do instead.
After that I got the job at GSK, where I was originally in the IT department. So I’ve always worked with computers and with figures and maths. I worked for them over in the States for a year and a half, then they moved me over to London. I was working in a role that kind of intersected digital and IT, and that’s where I met my Twizoo co-founder, John Talbot. He was part of a digital agency that did our social media monitoring and I was in-house and together we analysed the social media output to make the reports saying what people were saying about GSK on Twitter. It was never that interesting! We got along great and started discussing ideas of how we could put these skills into another business. We were tinkering around with the concept and our first idea was that it would be a review platform for all social media reviews, not just restaurants but what hotel should I stay in on my holiday to Paris and what book should I read next? All those opinions are already there on social media. We started with restaurants mainly because the data is so rich, people love taking pictures of their food, food porn is a thing, and it’s really popular area. We originally wanted it to be much broader with reviews of hotels, shops, movies – reviews of everything – but decided to narrow our focus to just restaurants to get it right.
Social media analysis is already a fairly major industry. Companies exist to tell other companies; ‘this is what people are saying about you online’ but nobody’s really taken that analysis and given it back to you and me in a way that helps people make a decision. We could have set up our own agency aimed at businesses, but we were really passionate about giving that information back to the customer because I personally don’t really trust reviews. There’s a lot of scepticism about whether the reviews are genuine or not, and more than that, often they are really old! I went on Trip Advisor recently because I was planning a trip and the review was five years old; I thought – this isn’t really relevant. People are sharing their experiences more naturally online on Twitter rather than review sites.
If I’m honest I had been interested in doing my own tech start-up all along. When I was in college I started a website – which no longer exists – called stylemystic.com, for young women to go online, fill in a profile about their personality and body type and then we would send them daily fashion tips and ideas about where to shop. I built the whole site myself from scratch, as I was taking a class at the time in web design and databases. I started it as a project for that site and then launched it for real. There was a discussion board on there and there were a bunch of girls on there talking in Russian, and I shut it down in the end because I didn’t really know what I was doing! But I suppose it was part of the building blocks of getting towards Twizoo, starting something myself, although I didn’t have any practical knowledge about running a business.
It’s funny, sometimes I think; this is way way harder than I thought it was going to be, but other days I think, well, this had to be easy because I don’t have an MBA in business or anything like that and look at how far we’ve gotten. It’s really trying to figure out what your strengths are and try and capitalise on that. For me, my co-founder’s background is in Computer Science, so he did a lot of the building of the app and my background in in maths and so I did a lot of the algorithms that analyse the data. So I would write that in mathematician’s terms in Excel or in a word document and I would hand it over to him and he would code it. It worked well as we were both playing to our strengths. I mean, I couldn’t write a book as I’m not good at writing, but I picked an area that uses the skills I do have. It was an advantage to be doing all the ground-work ourselves. I always advise people in tech start-ups to play to your strengths or, find a co-founder who will be in it 50/50 with you who has those skills. Some start-ups who outsource key elements of the work do work out, but it’s a huge risk to put into the business. Plus you will save a massive amount of money. If you were going start a restaurant you wouldn’t outsource the chef to someone you’d never met as it’s part of the core business. If you want to start an app, find someone who can code would be my advice!
There’s a lot of trial and error and re-doing things when you are developing your product, a lot of iterations to get it to a viable product. So we were doing all this development work while still working full time in our day job. We decided that if we could get financial backing we would quit our jobs and go for it full-time. We had built a really ugly version of the app, we had gotten some foodies and other core users together and showed it to them and said – what do you think, would you use it? And then we took that feedback and the data to investors. Once we had investors’ money we quit our jobs!
I remember that my co-founder celebrating both rounds of funding, because now we’ve had two, at the Hawksmoor, and thinking; ‘somebody else besides me and maybe my parents actually believes in this idea’ so it’s a big validation.
There’s a very good network of investors in London, there are people who are keen to invest in start-ups. It’s a business for a lot of people, who have seen how early investors in things like Facebook made tonnes and tons of money and they definitely see it as a business deal and are looking for a return on their investment.
In the time we were developing the beta version and even earlier versions we participated in an accelerator program and they teach you a lot of what you need to know to have a start-up and at the end you can present to a room of investors, that’s the kind of final stage of the program. There’s a lot of different models, some take equity, some don’t take equity, there are various ways. But the accelerator was really how we inserted ourselves into the start up ecosystem.
We had to do our sales pitch to the investors. We got better at it as we went along, but the first one I had to do, I was so nervous. I knew they were going to ask horrible questions, I thought I woudn’t know the answers, but then after I’d done a few I realised a lot of the questions were the same, I was able to anticipate what they were going to ask, and what resonated with investors was this statistic that for every visitor to a site like Trip Advisor, only 1% actually leave a review. And that’s very small, and there are actually seven opinions on Twitter for every one on a traditional review site. And then when we able to point out what a huge opportunity that represented. And how the review sites were spending a huge amount of money trying to persuade more people to leave reviews whereas people on Twitter need no convincing or persuading, they are already doing it. The data is already there so let’s just pull it in.
We spent that money on staff. Back then when we pitched, it was just me and John. Now we have seven people on the team. When we started we worked from the British Library – it was actually great, I miss those days a little bit. We used to go in a little reading room, it was quiet, nobody would disturb us, the Wifi was actually great, there’s even a café.
Now we have an office, with a sign on the wall that says Twizoo. The investment was in April 2014 so less than two years ago. We launched for real in July 2014, in London. Most recently we have been rolling out in cities all over the States, 12 in a row, in fact. 12 Cities in 12 weeks, as we are just pulling the data in from Twitter we can switch it onto a lot of different places.
The Twizoo app is free to download and our backers are actually fine with us not making money – I know that sounds a bit crazy to people who aren’t in the tech start-up world, but the way we work now is that we are testing out a lot of different business models to see what gets the most traction and eventually we will hit on the best way to bring in revenue. We have been testing out restaurant advertising, and you can actually book a table through the app. The third thing we could do it leverage out our data on restaurants out to third parties.
The past two years have been really busy. With all the different iterations of the app, tweaking and fixing bugs etc, it has been non-stop. But we are going up against these companies that are huge, and have millions and millions of pounds to spend. We have to move fast to see what works and to convince people to change from Googling ‘restaurants near me’ to using Twizoo instead.
In the world of apps its all about using the data, and the figures of who’s using the app and how they are using it, as well as a quite a bit of following your instinct, or your gut. You need the data but you need your gut as well and there has to be a balance of the two.
There were some things that we really wanted to have in the app; quirky features. We thought if it was gimmicky, people would switch – we continually thought; what is going to get people to switch from Top Table, or Yelp or whatever and we thought it would be quirky features. We had, for example ‘a drunk food finder’ that switched on after midnight! What we figured out in the end was that it’s the exact opposite. You want it as simple as possible to just get it to do one thing, but really well. Our thing is location – so restaurants near you, right now. We worked and worked on getting that as good as possible and scaled back a lot of the quirky ideas that weren’t really necessary. We can monitor the effects of each change really easily, looking at the data. We put it out there and then see what people do with it straight away.
In the future I want Twizoo to be <the> customer review platform on Twitter. I’m all in for Twizoo, this is it now, but further down the line I want people to go on it for movies, and not Rotten Tomatoes or Time Out or whatever. There’s more real-time, relevant opinions on Twitter than anywhere else and it’s a long-term vision but that’s what would be the goal for me.
If I had to come up with a motto that’s helped me keep going through all this time it would be; you have to be very thick skinned. Because the default answer that everyone gives you is ‘why would you do that?’, ‘Why would you leave a steady job to go off and start a restaurant app’, so you have to really believe in what you’re doing and just keep going. If I got down in the dumps every time someone said; ‘this is a crazy idea’ or ‘why would you leave your job’ or give me some scepticism about what I’m doing I would never get out of the bed in the morning!