Season 4 / Episode 3

Candice Ren – 173TECH

Co-founder of 173TECH

In this episode, Joyeeta is joined by Candice Ren, co-founder of 173TECH, an analytics consultancy that helps startups of all stages.

Candice talks about finding a North Star, a beacon which you can follow in order to grow and succeed in both your professional and personal life, and also discusses her passion for life-long learning, life-altering speeches, and how data can inspire significant changes to drive business growth.

Candice also covers her time with leading dating unicorns Badoo and Bumble, where she built and scaled a hugely successful analytics function before a $3 billion acquisition in 2019.


Moderator:  Hi, everyone. Welcome to Samsara again. We have a new guest on our podcast today. Let’s welcome data scientist, entrepreneur and a valuable member of our ecosystem, Candice Ren. Candice, welcome to Samsara.

Candice Ren:  Thank you, Joyeeta. Thank you very much for having me.

Moderator:  It’s wonderful to have you here, Candice. It’s not often that I host a data science show and I have another woman at the other end as well, so I’m quite excited to have you. Okay, so let’s get started first things first. Why don’t you tell us a little bit about yourself, Candice? What do you do? Where are you from? What’s your current role in the ecosystem, and what is your current company that you founded? What is it about? The whole story.

Candice Ren:  Sure, so currently I’m the founder of 173 Tech. We started end of last year. Prior to that I was the head of analytics for a data unicorn, I guess that’s what you call them, so Badoo, so it is a dating app, a group of them. It includes apps like Bumble, Badoo, Chappy, which is a gay dating app, and a few others, and was sold to Blackstone last year for £3 billion valuation. So, that’s pretty much the career highlight of mine. We started building analytics from scratch, having a full team looking at how data could help start-up growth, and that is how the idea of 173 came about. We decided to see how these things could help earlier stage start-ups, how can data be kickstarted in a very efficient kind of way, and also cost-effective? So, that’s the mission we have right now. In terms of myself, so I’ve moved around a lot, so I was born in Beijing in China, when I was eighteen moved to Australia, spent a few years there, and then moved again to Paris for my master’s, and then London six years ago, and the rest is history.

Moderator:  Wow, you’ve had a really exciting life and career, Candice. No wonder you have a well-rounded world view. So, tell me, I haven’t dived much into this world of data analytics and data science in the world of dating apps. Tell us a little bit about how you ended up being there, what’s so exciting and what sort of work do you really do broadly-speaking?

Candice Ren:  Yes, sure, so that’s a, I guess, blessing in disguise at the time. So, I started my career in broadly called, I guess, business intelligence, so that’s building dashboards, so finding out automation of reporting tools, and at that point I was thinking, you know, I’ve always loved telling stories with data, and I was using this tool called MacroStrategy. You could have a very long and sophisticated career within that particular segment, but I wanted to broaden a little bit, so Badoo came about, which was a product analyst role, which involved using a broader range of tools and being a bit more business-focused, so helping the app grow, finding out your user behavioural trends, things like why would people vote for someone in a certain way, how people are connected, what encourages or discourages them from continuing conversations, or using particular features in the app. So, that was particularly interesting for me at the time, and since I joined there we’ve gone through quite a few upscaling (inaudible 04.11), so beginning we were on Excel SQL and then we started doing more data science, so that involved Python and all sorts of different data science packages that you could use, and then productionising those algorithms, and at the end of the day it’s bringing people together basically, that’s what the app was meant to be. There are a lot of things we’ve tried and improved on, having failed a couple of times here and there, but at the end of the day it’s always focusing on what you want to do instead of artificial numbers.That’s one of the key things we’ve learnt earlier on, because for a dating app if you want to increase activity, what does that mean? Does it mean people voting more, but if they vote more but they’re not matching that’s actually obviously quite discouraging, and then we move to optimising towards matches, but then you find people match, but they don’t talk to each other. Again, that’s quite annoying for people, so a long way at the end we’ve decided to optimise towards meaningful chat, so that is something meaningful being you had a meaningful connection, so that is one thing, a user experience-driven approach to data science, I would say.

Moderator:  That’s incredible. I think that may potentially be all the things that, kind of, in some way resonate with other types of data science as well that I’ve been involved in. You’re absolutely right, you know. It’s not probably just about the numbers and seeing them, it’s what they really mean.

Candice Ren:  Yes.

Moderator:  Asking the right question is so important, it’s the first step in data science. The other day I was seeing these numbers which were about Slack versus Microsoft Teams users, and obviously Teams has surpassed Slack now by several million, and it may well be ahead of Slack, but for me those plain numbers meant nothing in terms of users. I was very interested in who’s actually using it. Is it just sign-ups? Is it like a bundle package that Microsoft gives to everybody, which is why this graph looks great? Are these real people who use it? Just having users is probably not enough. It’s what they do. That sort of a framework, I’ve tried to apply it to almost everything we do. We were looking at our own users, which are probably a little bit different from what you’ve done. They’re (mw 06.40) users, they’re probably not very B2C, but in the end for us also adoption is way more important than just simply seeing a large uptake, increase in number of users. So, I’m absolutely in agreement with you, long story short. So, Candice, what do you think is an important thing to do in order to arrive at this mind-set? Now, we all know in the world of data science that asking the right question is very important, so how can we go about cultivating this mind-set? What should be a practical way to arrive at that instead of flowing away with the numbers? How can we start by asking the right question? Is it by understanding the business first, or do you think it comes more from the data side? Where shall we start?

Candice Ren:  That’s a fantastic question. There is no right or wrong answer, but personally I would always start with your business. There are many times we are faced with the situation here, is five years of data, try to do something with it. I guess every data scientist or analyst out there has had that question being asked at very points in different businesses, and most of the time you’ll find, first of all, data quality is an issue, definition is another issue, so it’s not the most efficient way personally. I think if you have really well-structured clean data and you have a few years’ worth and wanting to do some exploratory analysis, that can be very interesting, but from the very beginning of the user journey, from a product journey, if you are building anything, in an app or a tool, so what you want to start with, so you talked about vanity metrics, that’s what I like to call them.

Moderator:  Yes, absolutely.

Candice Ren:  Yes, they’re in vain, they are useful to begin with, so looking at how you’re trending, so when you only have MVP and you want to just know are people even interested in using your product? That’s absolutely okay, but once you start scaling that becomes less so. So, we actually have a process of doing this. So, once you have daily active users or total active users, then you start looking at the intensity of users, so then you get to your point that you know your power users, so a power user is someone, for example, if you have a social app like a messaging app, for example, they will come back every day to your app, so they’re powerful, they’re always there and they’re likely to be very influential, they’ll be telling their friends about your app, they will be bringing more people in to you, having a very good ecosystem. So, those people, once you know them, then you can start doing some correlation analysis, looking at what made them a power user, and once you do know that, there are a lot of things, we talk about an ah-ha moment, it might have been a term invented by Facebook, I’m not entirely sure, but ah-ha moment is basically the moment, that ‘ding’, the light bulb moment, when people realise, (TC 10.00) ‘The reason I use this is because it helps me do A, B and C.’So, that’s your value proposition, that’s when people are hooked per se, so I think in very early days Facebook has, from their data, realised if people have more than seven connections they’re more likely to be retained as a long-term user. So, once you have that you can replace your vanity metric with what we like to call a north star metric, so instead of saying how many users you’ve acquired, you want to say how many people you’ve acquired that have added more than seven connections in the first week.

Moderator:  Yes, that’s amazing, I love what you said, replacing the vanity metrics with north star, like a guiding beacon, so you, kind of, start to solidify your business proposition using data, and then use that as a feedback loop to correct the very process of looking for things in the data, which is really a business first approach, and honestly, I think that I couldn’t agree more with you. We’ve seen it on our side within Gyana as well, that we do see a massive growth of users in this and that, but like you rightly mentioned, it doesn’t necessarily indicate success of the product unless we deep-dive into what sort of metrics in the product actually signify success, and how do we amplify that, and for us that’s been a couple of things. We’ve found that adoption, for us it’s not just the length of the session, that’s probably a binary thing. Once they exceed five minutes it, kind of, starts to get positive anyway. It’s zero or one, less than five minutes, they’re not finding value. What we’ve found is it’s the number of times that they log in within a period of not a week, but actually two weeks, because they work on data science projects, on our tool, and those are projects that you do for a little bit, you don’t necessarily just publish them right away, so you need a couple of weeks and you work on them. So, for us two weeks is the time period in which we check stickiness to see how they’re doing. Well, we’ve arrived at this much later.Initially, like everybody, we were just excited about, ‘Oh, they’re users and they log in everyday,’ but then we were, like, ‘Okay, does that even mean anything? What does it really mean for the health of the product?’ It’s still a journey, Candice. To be honest, we don’t know all the answers and we still think that there is a lot that we are yet to figure out. There are so many parameters that we don’t yet understand or track that could have valuable pointers towards better business solutions. So, how did you do that? Could you share with us some stories where there weren’t some parameters or things that you were looking at, but then eventually you found over the course of time that maybe adding that to your analysis could make a significantly better outcome, and how did you arrive at that? I don’t know, it could be luck, it could be a story, can you share something with us?

Candice Ren:  Yes, so there is no one formula, so it depends on what you’re looking for. If you’re looking for that one north star that might be a bit simpler because you know the key features that users are very interested in. Sometimes you can have surprises when looking at particular drop-offs, for example, so if you have a conversion flow you just realise this is an amazing feature you’ve released, but people are not using it. Why not? Without deeper data you probably come to the conclusion people are just not interested in this, which is quite unfortunate and you might kill off features that are a big winner, but it has certain bugs at certain conversion points, and this is where even tracking becomes really interesting and important, so in the app world that is when you have-, even web platforms are exactly the same, so that is your user journey, so how they’re converting from when they log in, they see a page, they click onto the next thing, so at which point you see a lot of people disappearing, so you will know that something is going wrong there, and from there you want to investigate more, there are different mechanisms you can follow, for example, breaking down by different demographic criteria, for example, for your client, what industry, or there could be stages of their, how would I say this, data savviness.There might be things that are not quite obvious to users, they might be missing certain points, so once you can zoom in to the problem, it might not be a problem once you zoom into it. Maybe there’s a massive jump at some point. If you can zoom into that you will learn easier what has actually attributed to that fail or success. So, data is important, but also knowing where to look at.

Moderator:  Perfect, I think that is great, and do you know what, which is one of my biggest weaknesses, our meetings where we do this thing where we ask the why, the question, because the question is so important, and leads to analysis in a completely different direction then if you had not asked those questions before, but like you mentioned, exploratory analysis could also be very strong. So, now I want to switch gears a little bit. I know there are quite a few people listening to Samsara right now, hearing you talk about all these exciting things you did, and a brilliant career that you’ve chopped out for yourself, and they’re thinking, ‘How can I do this?’ There are lots of women out there, not so many in data science, to be honest, but they’re thinking, ‘How can I make an entry, and how did you manage your career?’ so I’m very interested in knowing how you ended up with this. What did you study? Did it just happen or did you plan to have a career like this? How did it all lead to you being this data science (mw 16.12), this data science guru that you are? How did it happen?

Candice Ren:  So, it’s both the opportunity presented itself and also a bit of personal will, so I did a computer science degree back in school, and then I’ve always loved telling a story, and with a bit of a geeky side, numbers really facilitate that. I found you always have a much convincing argument where you have data to back you up, but really getting into the data science industry, that’s really thanks to my time at Badoo. Any start-up in general, I’m pretty sure it’s the same case at your company as well, there are just so many opportunities to learn. The two things I always followed as a north star for myself are you should always do what you like to do, and second, always learn more every day. So, with data science, so at Badoo when I started, that was after a period of time in marketing, so I dabbled in and out of luxury brand marketing for a bit, especially at the time when I was in Paris, and then when I started at Badoo technical skills was probably two years back, and just learning from your colleagues. I remember learning SQL from the guy sitting next to me who was fantastic, really skilled analyst, so I would get his code, read it, try to understand, try to make changes, and then just ask anything, if you don’t understand, and then try to get learning in. In this day and age you can learn without going to school or without going to a training programme, you can learn online. There are a lot of new types of data, academy, I guess hackathon, all those online platforms. You can just start. I did Python that way. I learned D3 JavaScript that way. So, yes, just really have time to learn and do that constantly, so planning in static time every week or even every day if you can.

Moderator:  Love that, it’s so true. Candice, I love that. I will tell you that I feel like I’ve spoken to someone who-, you’re really close to my heart now, okay? You’re just saying all the things I agree with, really. This constant learning and self-initiative is probably so important. It’s not just data science. I guess probably anything, and it just doesn’t land on your plate, it’s not a set path, you have to push through and figure it out, that’s what you did. You didn’t just sit down and take your degree and just continued, you kept learning, you kept pushing, and these days it’s so much easier than before. You can go to Coursera, you can go to Udemy. In this quarantine I ended up doing a few data science courses to refresh myself once again. I did one with Stanford, I did Andrew Ng’s machine learning all over again to see what he’s talking about, I did foundations with RStudio at Johns Hopkins, because I haven’t used RStudio so much lately, only moved on to other tools, so I thought what I’d be really doing, and I’m not going to use so much of it in my work, I do use a little bit of course for my own analysis in the business, but mostly my tech team does it. As the CEO I probably have to look at applied data, but it’s so much more empowering to know that I can use this so I understand my customers, I understand my tech team, and sometimes I use it directly for my own work, analysing (TC 20.00) Series A investment from Crunchbase, figuring out my own templates on, ‘Hey, what should be the next stage of the company?’It’s a really liberating feeling, but I have to tell you that some days when I’m doing it I do think, like, ‘God, am I a bit crazy? I have been around for a really long time and now I’m going back and redoing a foundation course to use it, so is this the kind of education I really need, or should I just focus on experience?’ Like you said, it’s not like that. We always have to keep pushing for educating ourselves and taking ourselves to the next step, and I think that you’re amazing for saying that, and I know a lot of women and men will probably be inspired by hearing from you, and they will try to do a few things on their own and keep learning as well. That brings me to a slightly personal question. Candice, tell us a little bit about who are the contributors to you being like this? Who are those people, whether it’s family, friends or teachers, I don’t know, to have this kind of attitude there must have been some sort of a support ecosystem, encouragement, inspiration maybe, I don’t know. Are there some people or someone you want to share with us who inspire you to be this person who’s learning, who’s pushing, who’s constantly setting an example?

Candice Ren:  Yes, there are just so many things. I think for me it’s more being in an environment, again, I draw back to my Badoo experience where I had a team that’s constantly learning, it’s not just me doing this, it’s really where you used to have a Friday session, so every Friday everybody will come and teach everybody else what they’ve learned in the past, and we all rotate, so the idea being if I learn something about D3 visualization I can come and tell someone else about it, and if someone is a marketing analytics expert they can tell me how that works, so having a very flat structure and also learning our our data environment, that really helps, surrounding yourself with smart people or driven people, but external inspiration, this is sounding a bit cliché now, but I think there was this speech Steve Jobs gave many years ago at the Stanford graduation ceremony, he talked about his own experience.

Moderator:  Yes, ‘stay hungry, stay foolish’.

Candice Ren:  There you go.

Moderator:  Yes, I loved that one. Oh my God, you are truly my unspoken best friend, okay? Absolutely, I loved that speech and it resonates so much. I probably honestly did not not wear shoes like he did for (mw 22.51) km, but I have to say it resonated with a lot of my struggles coming from a developing economy, and how I actually believed in the power of my dreams. Honestly, it inspired me so much through the years, and I cannot believe you’re saying the same. Wow.

Candice Ren:  Well, I guess he inspired a lot of people and especially given the industry we’re in, yes, that speech has been so long ago and I remember studying it in public speaking lessons, having heard it before. In terms of learning he mentioned about when he learned about type forms, so the design of different-,

Moderator:  Yes, calligraphy.

Candice Ren:  Yes, exactly, calligraphy, so when he learned that and he didn’t know how he was going to use it, he just wanted to learn that, and then obviously he invented the most beautiful font in the world, and contributed to Apple’s growth, so every time I hear that again or recite it, it gets me really excited.

Moderator:  I could not agree more with you, definitely true. I have just heard that speech so many times sitting in my room thinking about all the things you’re saying, so that’s amazing. See, this is the thing about people, they do this stuff, and I’d heard this famous thing, I forgot who wrote it, but it was about when you shine you give other people unconsciously the permission to do so as well.

Candice Ren:  Oh, beautiful.

Moderator:  Yes, it’s amazing, and I think that some day I will probably try to turn that into a big poster, yes. So, Candice, tell me a little bit about what do you do when you get some rare time?

Candice Ren:  My free time, when I’m not geeking out, so I’ve got a personal hobby of wine, so it actually can be quite geeky as well without a lot of people realising it, when you get into the technical side of things, so I’ve done some wine appreciation courses and I get together with people doing some blind tastings, so that’s probably how I unwind, with wine.

Moderator:  Perfect, unwind with wine. So, if anyone out there has a terrific data science project and they want to rope you in, you know what to do, people. Just send her a terrific box of wine. When all this madness is over, Candice, you and I will catch up with a glass of wine as well and see if we can discuss a bit more about data science, the world and our career. So, my last but not least question is for you to leave a message, okay? So, let’s imagine, well, not imagine, it is actually happening, lots of people are right now listening to you, there are stakeholders and various types of decision-making processes, some of them are business owners, some of them are investors, maybe some career enthusiasts, different types of people are right now listening to you, so you can potentially send them a message right now about the importance of data science, and whether we need it or not and how much we need it, and if you were to speak to them what would you say?

Candice Ren:  Yes, amazing, so data, there’s no doubt we need them. Everybody does. Data essentially, we can talk about data science, analytics, engineering, but essentially data is to help efficiency in our lives, so make our lives easier, we can make better decisions, we can avoid wasting our resources, so it’s not having a data strategy, it’s you must and you must start with it, and also, for the start-ups out there, if you’re doing a data strategy, one of the key things you might be faced with is immediate results versus long-term scalability. That is something, it is a very tricky balancing act. I would like to say don’t sacrifice your long-term ambition with your immediate gain.

Moderator:  Perfect, that’s beautiful, long-term ambition, immediate gain, know the difference and stick to it. Candice, I’ve really enjoyed speaking to you. You’re truly a remarkable woman. You’ve done amazingly in a field and a discipline that is quite hard, and you have led so many companies to huge success, and now you have your own firm, 173 Tech, which I’m pretty sure will go on to do amazing things. It’s great to have you in our city of London, and I’m so inspired and excited by you. I do want to end on the quotation that I mentioned a little bit before about the permission, and I think that you will really enjoy it, because you know what? You reminded me of this quote so much, so I’m just going to read it out. It’s written by Marianne Williamson and it goes like this. ‘Our deepest fear is not that we are inadequate. Our deepest fear is that we are powerful beyond measure. It is our light, not our darkness, that most frightens us. We ask ourselves, “Who am I to be brilliant, gorgeous, talented, fabulous?” Actually, who are you not to be? You are a child of God. Your playing small does not serve the world. There is nothing enlightened about shrinking, so that other people won’t feel insecure around you. We are all meant to shine, as children do. We were born to make manifest the glory of God that is within us. It is not just in some of us, it is in everyone, and as we let our own light shine we unconsciously give other people permission to do the same, and as we are liberated from our own fear, our presence automatically liberates others’. Thank you so much for speaking to us today. I will definitely catch up with you when this madness is over, and we will get some of your famous wine.

Candice Ren:  Fantastic, that is beautiful, and thank you so much for your time and this podcast. I really enjoyed speaking with you too.

Moderator:  Thanks, everyone, for tuning in today. See you next week.