How can you kickstart a product journey? Really understand your users? And decide on the right time to hire a CPO?
We got the chance to sit down with Tugce Bulut, Founder of Streetbees, as she took us through her fascinating success story.
We covered so many important topics; from mental health and support networks to startup realities and how to build your network. Here, we’ve got some of the key highlights from the session, but if you want to listen to the full interview, simply click below. 👇
Q: Tell us about Streetbees pre and post-COVID. Did you have to evolve the way you work?
A: Streetbees was already quite a decentralized organization. After all, we gather data and real-life observations from all around the world.
We currently have people on the ground in China helping with community building, we have people in the Philippines and Vietnam who help us with data classification and tagging, etc. So we are very used to working remotely.
But one thing that you take for granted is team spirit. We had an office in London, one in Lisbon, and people would get together. They’d have drinks, workshops, and sessions and that means so much, especially in a higher growth company that you can just jump into a room and draw things on a whiteboard.
All of a sudden, last year, we all had to work from home, and we haven't been back to the office since then. Although our work is completely doable, functionally, with just a laptop really, wherever you are, we all miss the team side of things.
Q: What was the pitch that you gave to potential investors for your Series-B? Did COVID play into that?
A: To be honest actually COVID didn't play so much of a role in that, it's an interesting story because we continued to grow very strongly before COVID and after. We were in a very cash-strong position as well because our unit economics is very strong.
And we weren't really thinking about raising and even when we thought about it, a lot of people were telling us that we should wait until the whole COVID thing had cleared up.
We do have regular meetings with many venture capital firms to build relationships. I always tell people choosing an investor, especially a lead investor, is like marriage. You better date and flirt for a long time until you make your decision because there's no going back really. Divorce might be easier than trying to part ways with an investor who is on your cap table.
But jokes aside, it's a very important relationship that you need to build over time. So we were always keeping in touch and a couple of the key funders we were talking to came forward and said you're growing extremely fast and with more resources, you can continue to hire more people in technology, increase your market coverage, increase your salesforce, and that will even help the company grow faster.
That made a lot of sense to us, so we started conversations with a few chosen partners. Then we made the decision to go forward because frankly, waiting for another six or nine months wouldn't make any difference from our perspective.
What the pitch is and why the investment community has always been very excited about Streetbees is that there are lots of interesting technology companies coming up. But it's very rare that there's one with the potential of creating a multibillion-dollar platform with deep proprietary technology. It's even rarer for that to happen in Europe. So I think we fit a very small group of companies, working on a problem that's going to be worth billions to solve.
It's done with a defensible technology with barriers to entry. The simplest way to think about Streebees is to think about the experience of the internet before Google. You'd have to remember the websites and we used to use bookmarks and actual Excel sheets to remember sites. What Google did is index, classify, tag every single thing that exists on the internet so that when you need it, you just go to a search box, and boom, you find it.
Streetbees is doing exactly the same thing for offline life moments, Google has done a fantastic job on the online side but actually, 80-90% of our lives that are still happening offline, are not available to search through. It's all in your brain basically.
So what we're doing is using the app that's with you all the time, you turn your real-life moments into a data point. As a first layer, we gather millions and millions of real-life observations - when you're eating, drinking, socializing, going to exercise, driving, whatever it is. Then we keep them in a data warehouse.
Then there's a second layer of the Streetbees technology, where we have our proprietary deep neural networks that interpret that mega-scale data we have in the data warehouse, and on top of that sits a knowledge graph that turns all that interpreted unstructured data into business intelligence to answer specific business questions.
So in the same way Google made the online world searchable, we’re making the entire offline world a searchable data source for consumer companies, for anyone who wants or needs to understand what people want.
This can vary from, for example, looking at the exact ice cream you're really enjoying - how do you talk about the texture of that ice cream? This is something you can search through in our database.
How people's anxiety during COVID is influencing Consumer Confidence Index, economic growth in that particular country or in a particular category is also something you can search through on the Streetbees platform.
Q: When did the understanding of data come into your product journey? Did it evolve as you found your product-market fit?
A: We didn’t actually start with a product idea. We started with a problem because I was actually experiencing that problem in my previous job as a strategy consultant. I was on the buy-side of data and we were working with massive companies from Hersheys to JustEat, for example.
We always need to understand what consumers want, what they currently do, and where the needs are that we can address? And it was very interesting to me that most data sets, like credit card transactions, point of sales data, macroeconomic data, you get time series, are all super important in predictions and modeling.
When it comes to the most important data point, which is the consumer experience; what they want, what they need, and how it's changing, you have to write a quick survey with multiple-choice options. There's a spot analysis, there’s no time series and you need to rely on and trust people know, as they can articulate what they need for you.
If I asked you, for example, when was the last time you ate McDonald's? Most likely, you're not going to be giving us the true answer. Not because you're lying because you can't remember, let alone remember your emotions in that very moment.
What was your exact context? Who was with you? What time of the day was it? You're just not going to be able to remember any of this. Unfortunately, surveys are the worst way of finding out about what people want. We wanted to create this platform where the real-life moments are happening, we can just capture it as an event happens and then turn it into time series.
We hadn’t even thought about the product at that stage in terms of how we’re going to actually collect that data? How is it going to be analyzed? How is it going to be visualized? That all changed step by step later, as we onboarded customers.
Q: What gave you the confidence you could solve this problem?
A: I think to me, it's more about is someone willing to pay for it? Because if there is enough money to be thrown at it, you will solve the problem. It's just a matter of is there a market?
And it was so obvious to me that there is a market, there is demand, there are not only millions, billions of dollars being spent, that's enough to be able to build the right technology on it. The two factors, though, that were already known to me, which I knew is going to make this possible is that, if there is a willingness to pay, what are the other two blockers that can happen?
The first one is that the users might not want to share. So that's something that you really need to think about and have access to that information. I was in a very lucky place with my background, when I was doing my Ph.D. and working with World Bank I used to do ethnographic research on the ground one-on-one interviews, my topic was on poverty alleviation.
We were basically speaking to families who are living under $1 a day and listening intensely, in-depth, how they describe their lives. It was really good learning for me, this is many years ago, it was good learning for me how much people actually enjoy sharing, as long as you're listening in the right way and you're offering them solutions.
So we knew that it was an inherent human trait wanting to share as long as you are treating their data with privacy, with respect, and really giving them back something in return for sharing.
This is what we thought that we are going to build the world's first fair data exchange platform where people are going to share, first of all, anonymously, their names are completely detached from the information they're providing so it can never go back and upset them in any way.
And they will be compensated for their contributions. Now, there are a lot of businesses that are using data as a sub-product from Google's and Facebook's to dozens and dozens of other businesses but the user is not in control of how their data is being used for what purpose, and they are not getting compensated for that.
So we wanted to change both, we wanted to make sure that the user is always in control. So you share what you want to share, and you want to share and you know exactly what are we going to do with that data? And secondly, you get basically fairly compensated for your contributions on that data as well.
I said at the beginning, there are two potential blockers. So that will be one. The other one is well, is it technologically just an impossible problem to solve? I strongly believe there is no such thing it's just going to be a question of how many years are we talking about? And how much cash do you have to invest in it?
What I didn't follow in my previous job was that natural language processing was moving quite fast and it was allowing us to already do topic modeling, for example, based on unstructured data.
There was enough early science to have faith that not only us, but a lot of other players, like Open AI, could continue to work on this problem collectively. Of course, we can solve the problem of understanding what people want from unstructured data that they shared with us in the form of a conversation.
Q: How did you decide it was worth delivering this product in a repeatable, scalable way?
A: Interestingly, that was definitely a lightbulb moment where I got quite tired of traditional solutions in the market. There was a moment I spoke to one of the traditional players, I said to them, "Listen, all I want you to do is build a mobile app that gathers the data in real-time, and all you need is a couple of NLP people who's just gonna analyze it, why are you guys just not doing that?"
I was quite upset they weren't even considering it. Because we had the willingness to pay, we were going to pay for it, if they just did it. And that night, I decided I'm just gonna do it. Because I felt that there is a buyer for the data, there's a seller for the data, and no one is making the market.
So the next day, I told my boss "I'm really sorry, I love my job, but I'm gonna quit because I really want to create this platform". He thought I'm just having a bad week and I'm thinking of quitting. But then after a few times listening to me, he actually became the first investor in the business.
Q: So what was your life like? Did you have all the support channels you needed? Was this scary?
A: Yeah, it was an absolute nightmare. I think the gist of the story is that if you really want to do something, you will make it work. There is no reason to think too much.
I think for some people, maybe that comes a little bit easier, in the sense that, once I made the decision, my thinking was, what's the worst thing that can happen? I will give this a go for a year if it really doesn't work, if we don't reach a minimum revenue, I'm going to go and get a job, because I can't continue to spend time on something that doesn't make money.
But we have been very lucky, very successful over the last six years. We did hit all those revenue targets I made sure to continue with the business, but even then, it was extremely painful and difficult for sure. Just to give you some examples, I quit my job, and then I said to myself, "I didn't think this through, how am I going to actually pay rent in the next couple of months?"
I had to move out of my flat in central London, and then started living with a friend for about four months. That saved me a lot of money. We had a very small team, and frankly, I was working probably 20 hour days so there wasn't a lot of time to go out and do anything else anyway.
It was very hard but to me, it was all about - if it doesn't work, it's just one year. And there is no point in keeping on doubting and questioning yourself. Of course, you're gonna have some awful days, weeks, sometimes months that are very difficult. But I wouldn't go back and question, did I make a mistake?
Now I'm committed for a year, I made the decision. I know my milestone, I need to hit it. So let's not avoid creating drama about every day, like constantly asking, is this working? Is this not working? You will know in a year if it's working or not. Until then just keep working hard basically.
It can be very small targets. my target for the first year was only 200K. That's why I said, the business needs to make $200k in revenue in the first year. Because that's enough of an indicator to show someone sees value.
We ended up making a lot more than that. We ended up registering clients like Coca-Cola and Unilever in the first year. So by the end of year one, there was absolutely no question mark in my mind about demand.
Q: When did you decide the right time to hire a CPO? Did you feel as though you were handing over the evolution of that problem solving to somebody else?
A: No, I couldn't be happier. It's a function that needs its own owner. In the beginning, as a founder, you also do sales, when I hired my Chief Revenue Officer, I was the happiest person on earth, that someone is actually gonna take care of it now. And instead, you focus on the future of the business and the three-year future.
As a CEO your job evolves a lot year on year as the company grows. Now, my main job is to make sure that we always stay ahead of the competition and what I'm working on right now doesn't have pretty much any impact on the next six months or even a year, we are already working on the next three years.
Again, part of your job is to align all these different teams so that marketing, sales, product, are all working towards a single goal. You can't do that and also be the Chief Product Officer who needs to go super deep into product unless you're Elon Musk, somehow he seems to manage it.
Or go into the depth of sales and give really good feedback to every single salesperson in the team, etc, you wouldn't be doing a great job at it, even if you tried. The other thing is that a lot of founders don't come from the bottom up on that function.
So you actually skipped a lot of the steps and important knowledge, which the person you are bringing in has, so they're gonna do a better job anyway, in terms of developing the product.
However, that doesn't exclude the involvement of a founder in product vision. You hire your team to help you achieve that vision. But the vision still comes from the founder.
Q: Where do you see the challenges ahead for making sure you stay relevant?
A: That's exactly the reason why you hire people like Chief Product Officers because it needs focused attention, otherwise, you can start falling behind. For us, it's all about leveraging AI technology for personalized conversations on the user/consumer side.
Why? Because every personal experience is so unique. And if you think about it, the whole world has been about trying to fit us into a single box so that they can service us a little bit more easily. But that actually doesn't work, because that box doesn't fit anyone in the end.
This is the beauty of the technology that's available to us today. We don't have to do that anymore because there are cost-effective ways of giving everyone what they want.
If you think about that, you come from a research background, and the surveys were all about asking the same five questions everyone, and then five answer options are already written for everyone, and I expect the entire world to fit into those five groups, for example. But the world doesn't work that way.
So what we're now working on is that we ask you something like, how are you feeling right now? And you answer to us in open text, in your own words, maybe a whole paragraph describing what's going on and what's happening in your mind. That was already the first step that we took six years ago, let's free people from multiple-choice questions.
But there was one more step for perfect personalization. Ideally, we should understand what you are telling us on the fly, and change the next question we have in the conversation. And that's the really groundbreaking change we are going to be bringing this year, that we are moving our AI to the device.
We use a lot of our technology in terms of neural networks and topic modeling with NLP on the server-side once the data is collected. But actually, we now want to do this on the fly. So you tell us let's say that you used to live with your sister, she now moved to the US, you haven't seen her in ages and on the fly, we can detect there's a sense of loneliness there.
But you're not so sure you didn't really use the word so we ask a follow-up question, how does that make you feel? And then you maybe say actually, I'm really feeling a bit lonely. Then we ask another question, what makes you say that? Or what kind of actions does that lead to etc.
We literally can automate a real human-to-human type of conversation happening to get deeper and more meaningful information from you while giving you a personalized experience so you don't feel like we are just asking you some framework questions that are not really relevant for you. That's exactly how you stay relevant to the user.
Looking for more insights on all things product leadership? Check out the CPO Space and get inspired. 👇