In this episode, we talk to Mihajlo Popesku and learn how Qudo is looking to revolutionise marketing and consumer research without compromising user trust and privacy.
Episode Links
Connect with Mihajlo on LinkedIn
Episode Transcript
Matthew Todd
Hi. My name is Matthew Todd and welcome to Inside the ScaleUp. This is the podcast for founders executives in tech looking to make an impact and learn from their peers within the tech business. We lift the lid on tech businesses, interviewing leaders and following their journey from startup to scale up and beyond covering everything from developing product market fit, funding and fundraising models to value proposition structure and growth marketing. We learn from their journey so that you can understand how they really work, the failures, the successes, the lessons along the way, so that you can take their learnings and apply them within your own startup or scale up and join the ever-growing list of high growth UK SaaS businesses
Hey and welcome back to the podcast really pleased today to be joined by Mihajlo Popesku from Qudo. Great to have you on the podcast today.
Mihajlo Popesku
Thank you so much, Matt. It’s my pleasure to be here.
Matthew Todd
No worries, I’m looking forward to hearing a bit more about the AI-based startup that you’re involved with, as well as a bit about your background that led you to that point. Do you want to to kick things off by giving us a bit of a brief introduction to yourself, your background and what it is that you are up to now.
Mihajlo Popesku
I’m a Marketing scientist. I did my masters, a PhD at University of Nottingham, where I studied value cooperation. And I was applying some advanced advanced quantitative technique to understand complexities about how consumers co create value. I always had this intention to join industry and bring the best from science, and apply the scientific knowledge. When I was doing my PhD and master studies, I had a lot of industry related projects. And I really enjoyed those. And I was thinking, how could I have more of these projects without actually sacrificing too much of my time, because, as you know, as a consumer researcher, you are pretty much applied, it’s essentially an agency model where you, you know, you design the research, you go into field work, you collect the data, crunch, the data, you know, find insights, and then present some insights and recommendations to your client.
Thinking back then, this was in 2014 2015, I was thinking, one of the most consuming parts of consumer research is actually me trying to understand one particular domain and then having to design a questionnaire or an instrument. This will typically take me like, a couple of weeks. Whereas the rest of the process, like, you know, collecting the data, and crunching the data and stuff like that, you had off the shelf off the shelf software. And you know, this was pretty much faster once the data was in yet. But that part before data coming in was painful and tedious.
So I was thinking, what is it that I could do to boost my productivity? And back then I was thinking, if I had this tool that would generate questionaires for me, or that would help me understand one particular domain, I could probably be way more productive, and I could do much more, my multiplier, my income multiplier would would increase significantly.
So back then I was thinking, if I had this platform that will leverage crowd wisdom, or if that would crowd source questions, or that would bring pools of experts that would quickly suggested what are the questions that I have to ask to cover a domain? That will be quick. And one of the ideas was to have this question bank where experts will be depositing questions. And whenever, let’s say a recommendation engine leverages that question, the author gets a compensation.
So in a way, your question or a questionnaire becomes your asset that is generating new passive income. I tried pitching this to the University of Nottingham back then, and this idea didn’t didn’t land well. And a couple of years down the line, I was a university professor for a while. And then I joined Cambridge Analytica, I met a couple of interesting people there. And then later on, we decided to do something of our own, and big ideas, you know, converged.
The idea of research being used for targeting and activation and then the AI being used to automate questionnaire generation, cross pollinated my original idea and this is how we started Qudo.
Matthew Todd
In terms of marketing insights, and what you’re trying to achieve then, you know, why is that an advantage to your customers to consumers as well, you know, why is that a better approach than the the targeting that these ad platforms already provide?
Mihajlo Popesku
If I may just add one thing so my current research is specializing in understanding or discovering different audiences or segments within general population, so that brands can better engage with them. The way how we design our research, we design our research in such a way to enable bridge between insights and activation. This means the way how we ask questions is allowing us to target people on social media without knowing anything personal about them without knowing their phone numbers, emails, or you know, a name, surnames, etc.
So we are looking at affinities. And then those are things that are, let’s say, very specific to certain group are then used to build audiences in Facebook or Google so that you can serve certain content to a group that will resonate well. That’s the advantage, you know, the existing model out there is, if you if you want to understand different segments, your traditional research agencies will give you a report or a dashboard. But then there is nothing telling you how you actually bring this to life, how you actually speak to the group that was discovered by that particular research, and we give you a technology that is crossing that bridge in a GDPR compliant manner using zero party data.
Matthew Todd
That’s really interesting. It is that gap, as you say, between that research to identify this is the audience, these are their interest behaviors and everything else that are going to resonate. Now you’re bridging that gap. And this is how you then get in front of an audience that wants to hear your message.
Mihajlo Popesku
Yeah, absolutely. And one more point, you know, big companies who own loads of first party data, first party data is everything that company or a brand knows about you as an individual. We’ve seen recently, and major research companies such as Gartner report that companies knowing way too much about you is creeping the consumers, like if you speak to someone, and then you show that you know more than three or four pieces of personal information, this is way too creepy, and it deters customers and consumers from engaging with your brand.
So we are giving brands, something that is completely void of any first party data. So you can speak to a group in such a way that you keep their privacy and yet show that you are familiar with their behaviors, preferences, and affinities.
Matthew Todd
I certainly see a lot of people are put off by that degree of over personalization having access to data that, you know, as a consumer, you, you didn’t naturally expect them to have. It’s obviously based on on, you know, tracking a lot of information and sharing of data as well as that first party data. I think, you know, we’ve all had the case where we open up Facebook, and all of a sudden, suddenly, we were just talking to a friend about has suddenly appeared in our feed.
Mihajlo Popesku
Yes, in psychology, this is called the recency effect. Because if you talk about something that is important to you, you’ve been actively thinking about this particular topic, and you are then more prone to notice these things around you. Maybe you already did the research on that. So you know, those kinds of stuff appearing on Facebook or Google, I wouldn’t say they are necessarily listening to you, it is just that what you do, what you search for is actually indicating that your next thing that you’re looking for, is going to be suggested by those engines.
I think the algorithms are very advanced right now. So, you know, the algorithms can now almost already predict what’s your next step going to be. So as I say, like, this is now creeping us out in a way, but in the same time, it gives some convenience of life, you know, makes me you know, makes looking for this new product that we’re going to buy or things that we’re going to download or watch, you know, more easily accessible to us.
So, you know, it’s not I wouldn’t paint necessarily this picture black, you know, there is a shade of gray. I will say those platforms know way too much about us, but in the same time, you know, they might simplify our life, but, but also there is, again, a negative effect that, you know, they might keep us in this like echo chamber, you know, and feeding us the stuff that maybe are not necessarily new for us.
Matthew Todd
There’s definitely two sides to it. I think. No one inherently minds being shown different types of content and adverts as well. As long as it genuinely does resonate with our interests. I think we’d all rather have that than something which is of completely no use or no value of no interest to us whatsoever, right. So there’s there’s definitely two elements to it, but as you say, sometimes it can feel a little bit creepy and that’s why well, brands probably don’t have the right impression.
Mihajlo Popesku
Yes, definitely. As I said, it’s been noticed that companies knowing way too much about us is definitely creeping out our consumer. So I would say, hyper customization and hyper personalization is probably going to backfire and has already started that backfiring big brands.
Matthew Todd
Is that also an element as well of those third party platforms, be it Facebook, Google, whoever, if they are the only way for you to identify and get in front of your audience that for many brands and businesses, as they start to scale, there can be an over reliance on those platforms such that if for whatever reason, their access to those suddenly is removed, or the algorithm there changes in some way that actually they don’t have any valuable insights of their own, that they’re able to leverage and make use of.
Mihajlo Popesku
I will agree with you, I think, you know, obviously, there is a risk of over-dependence on certain third party platforms. But when we talk about Facebook, and Google, these are currently, you know, the biggest advertising platforms out there, you know, Facebook, in our eyes, Facebook is not a social media, it’s an advertising platform, because this is their primary business model. They’re essentially making money by monetizing on their users. You know, in a way we as Facebook users, we don’t pay for using the platform, but yet the company has to live off something. This is actually selling us to brands so that we use audiences on which they they monetize, I think this is like a common place.
Speaking of that, the way how we are building platform, we are building platform to start with Facebook and Google, the integration with them, is important to us. Because as I said, you can tap into the biggest audiences. Of course, there is the threat of having certain audiences and being like over targeted or oversaturated by certain, you know, advertisements, but at the same time, we are working hard to become a channel agnostic platform, so that you can serve a certain message to relevant all the audiences in the channel that is relevant to them.
We’ve seen now the rise of Tik Tok, you know, you have so many other platforms, if you want to talk to let’s say business audiences, you have LinkedIn, if you want to talk to more to audiences who are, for example, more into debating with others, you have, for example, platforms such as Quora, or platforms, for example, such as Reddit.
But these are more more more of a niche platform. So in a way we want to become channel agnostic, and Facebook and Google for us were like the logical first steps in validating the products.
Matthew Todd
I can see why that that makes a lot of sense from you know, how you approach that initial validation, and then your, your vision for being channel agnostic, as well. So when it came to Qudo then to launching that to proving out the the concept behind it, what did that early journey look like?
Mihajlo Popesku
That’s a good question. We are still currently in some kind of stealth or semi stealth mode. And for us to validate the the effects or the impact of our product, or the the value that we’re delivering, we had a number of close beta, beta users. We had users from banking, some loyalty cards, some leisure and travel brands. And one of the things that we did, we did, for example, for a bank, one major central European bank was launching a online instant cash flow. They wanted to test how the platform like like Qudo perform against their traditional way of doing, let’s say, you know, social media engagement and conversion.
What happened is in parallel, we ran two studies, one study was informed by our research, our segmentation and our targeting criteria, and other study was there, you know, I will say blanket campaign. When we compare that for a month of running those both campaigns in parallel, their campaign conversion of 25% in our, in our campaign had three times almost three times better results.
So we managed to convert 72 out of 100 loan applicants, which means we managed to get way more people being approved for a loan, other marketing performance, such as you know, page cost per page, cost per click, etc, etc. Typically, you know, those other marketing KPIs were between to 50 and 500%, above and beyond what they had.
So in that respect the product was validated, we have a number of other case studies shown that like, if you really understand your audiences, if you understand what are their needs, how to speak to them, and what are their affinities, then you can definitely outperform any any guesswork that that, you know, marketers typically tend to resort to when they have no data.
Matthew Todd
So you’re able to improve those typical marketing metrics, the costs per click and the cost to, to get that, you know, lead prospect data, but then you’re also able to optimize from the actual sales side, as well, as you say, if you’re getting a much, much higher percentage of those loans, applications approved, and that’s got to be a big operational efficiency for them as well, hasn’t it?
Mihajlo Popesku
I mean, we managed to improve the key metric, and that’s RAS, return on advertising spending.
Matthew Todd
In terms of the, the customers that you’ve worked with, so far, then is that a very similar kind of campaign based model that you work on?
Mihajlo Popesku
Our target markets are UK and USA, and we work with commercial commercial entities, or consumer brands. But our clients are coming from different parts of the world, we have clients from, from Eastern Europe, from, from the UK, from North Africa, again, different commercial entities.
Our stepping stone into the market will be two studies, we will do financial services study in the UK and USA. And we’re going to do a general consumer study or general consumer profiling in both UK and USA. For the next six to 12 months, we’re planning once we launch and we’re going to launch by the end of this year, we’re planning to, you know, approach the market with a freemium model where, where everybody and anybody will be able to see the segments and then use those segments for activation.
Then we’re going to listen to our communities see what was well received, in what was liked in our product, what was not so liked, and then we’re going to, you know, improve, refine, and in one of our next iterations that we’re going to see, we’re going to see, like, what kind of additional features can then be paid for, essentially, this first version coming to the market will be freemium.
Matthew Todd
For that, you know, more general consumer use case and there must be, you know, a massive, massive amount of analysis you you have to be running in order to enable something like this.
Mihajlo Popesku
Yeah, definitely, we plan to run largely representative samples collect large representative samples, every quarter, I’m talking about 4000 respondents in UK and USA respectively. We’re going to understand we’re going to try to understand different things about how people in general approach consumption was their relationship with brands with spending, you know, what’s their, you know, mindset in general, you know, how they make decisions about certain purchase, what they value in different companies, brands and offers.
So, then we can give you like generic top level segments were talking similarly about between seven and 10 segments, and we are going to follow important trends, trends such as well being or sustainable consumption, privacy, consumer rights, so to see how people perceive all these are the important aspects of their lives. So that once we will come up and once we bring to life, those those segment personas, different brands will then be able to select and say, okay, well out of this, this 10 segments, I think those two or three are relevant, so I can then use cue to engage with those audiences, and, you know, maybe build awareness about my brand or, you know, build sales, etc, etc.
Matthew Todd
That makes a lot of sense, I can see how you can use that kind of affinity information to better get in front of the the customers that are likely to buy into your brand, and brand values, as well as you know, want to need the product that you’ve got to offer.
Mihajlo Popesku
You don’t have to know first party consumer data to approach certain relevant audiences. You need to understand affinities of certain groups, so that you can engage with them in a GDPR compliant manner.
Mihajlo Popesku
It is very important to understand consumers, lifestyles, their mindsets, their affinities, because affinity is a nice targeting and heavy targeting criteria that was that is offered by both Facebook and Google. This is what what essentially we are arguing you don’t you don’t have to know first party consumer data to approach certain relevant audiences you need to understand affinities of certain groups, so that you can engage with them in a GDPR compliant manner where there is a clean handshake, you know, no one’s data is involved. This is just like a broad level understanding of how different groups approach come assumption, how they go about engaging with brands, what is important in their lives? And what are their affinities? So yes, there is alternative, essentially to using first party data, and we hope to be champions of that new philosophy.
Matthew Todd
Absolutely, I think restrictions are only going to get tougher. I think, you know, with different blocking mechanisms in browsers and everything else, I think, yeah, people are trying to, you know, limit so much of that exposure of, you know, the personal identifiable data. And I think, as you said before, I think consumers are starting to push back a little bit against that, as well and realize how much your personal information some of these platforms really do have.
Mihajlo Popesku
The truth is first party data, it always belongs to consumer. It doesn’t belong to a company. Many companies, despite the fact that it belongs to consumer, they’re monetizing on that data, not passing any of that value down to the consumer. Our philosophy is the following. All our data is coming from anonymous surveys, we incentivize people, whatever they say, in the survey was shared with us on a voluntary basis, we have very clear terms and conditions, we explain how the survey is going to be used.
As I said, everything is voluntary. We ask no personal information. We just ask about your lifestyle, your views of the world, your attitudes, and we use that to build. We also ask about your past behavior. Then we use that to build different consumer groups that are then used by brands to engage with audiences who look like you. Not necessarily you personally.
Matthew Todd
That’s a really important difference. From what you’re describing, it certainly sounds like what you’re trying to build with Qudo is very, very different type of marketing technology than is currently available.
Mihajlo Popesku
I would say more than that, we are advocate advocating a new philosophy, as I said, you know, zero party data, everything GDPR compliance, your privacy is protected, we don’t, we don’t go and then merge the data set that we have with something else that someone else might have.
This is very common now in the industry, we speak to different technology providers, they buy financial data and they start cross pollinating all these datasets with us, you know, we just save the data that we asked you from the survey, we don’t know who you are, we can’t merge anything of this insights with any of your private data. So for us, the philosophy is keep privacy protected, okay.
Second thing is deliver value to both brands and consumers by essentially helping consumers be advocate of the consumers and their shifting needs, and also help brands to engage to the relevant audiences where they can cater to serve the need or fix certain problem. So I think the our philosophy is actually been translated into this new piece of technology, and we call Qudo precision activation engine. We believe we are building a first of kind product that is going to be launched by the end of this year.
Matthew Todd
I can see how that philosophy is embedded through everything that you’ve just described, about the way the the product works. As well as the way that you, you say, you’re hoping to take it forward, as well opening it up to more general consumers.
In terms of, you know, how far Qudo has come down from that, that starting point, I’d love to hear what some of the challenges are, that you’ve experienced when trying to build something, you know, which is very, very new and very, very different to what exists at the moment.
Mihajlo Popesku
So Matt to be very frank with you our challenge was not in how we build technology, our challenge, our key challenge was finding the right talents. Building the team here in London, where we had like a cutthroat competition was a very, very difficult task. I mean, we really take pride at the fact that you know, we are organization that has like a very good well-being program and our people are engaged and happy.
But you know, when you are a startup that has very little on your website, it is difficult to attract the right talent. Even though we offer like good package, and our people are super happy. Its super difficult to find the right the right engineers super difficult to find the right data scientists super difficult to find the right you know, social scientists. So it took us way more than then what we anticipated. Brexit didn’t help by the way, because like we’ve we’ve seen that there is like a limited supply of good talent here in London.
So that was one challenge. The second challenge is our vision was way too broad and very ambitious. So like prioritizing what you build first. So for example, because I’m researcher, I wanted to build this survey hosting platform that had artificial intelligence, questionnaire design embedded. But then my marketeers said like, well, actually, you know, let’s rely on a third party solution. Then let’s build everything else. Let’s streamline, let’s automate the analysis. Let’s build this bridge between insight and activation. Let’s do something lighter, that we can go to the market faster with, and then we can validate. And then if, if it works, and we’re going to then look at the deep data and build all the AI and stuff like that.
So for now, we decided to, you know, build a bridge between inside of activation, and it seems to be working well. Going forward, we are going to minimize our technological dependencies on third party, third party platforms, such as server hosting platforms, or panel aggregators, etc, etc.
Matthew Todd
I can see how it could be difficult to find that kind of talent as well as navigate that decision of what it is that you’re building and the order to build that in as well. Because what you’re what you’re talking about is a it’s not an insignificant problem that you’re you’re trying to solve right is a pretty complex problem that needs quite a lot of data. So imagine the, the roadmap for what you want to build could be could be massive.
Mihajlo Popesku
Yeah, by all means, like, our idea is to have the full marketing communications, value chain, streamline and automated when I say marketing communications value chain, I’m talking about those communications, powered by data. We’re talking about research data survey data. In that respect, if you’re starting from research, you need survey hosting platform, you need an algorithm that is able to generate questionnaires.
I mean, the ultimate idea for us is to have this SaaS and to give it to your average market here or your average researcher so that they can leverage value from that. In a way, the idea for us is to minimize the pressures on experts. When I say experts, for example, research expert, data scientist, and digital marketing. We want to like, as I said, streamline and stitch all these processes together by expert quality research, expert quality data science and analytics, and expert quality digital marketing and activation.
Currently, if you’re a market here, and if you want to engage in data driven data driven digital activation, you have to play with a number of tools, or a number of themes. This is a massive complexity and a massive headache. No one has time to do that. But if you have a piece of software, in a way, an ecosystem that is essentially helping you to seamlessly progress from research to analysis, from analysis to activation, you have then first of all synergetic effect, you’re not losing any signal between important steps in the chain.
Finally, you know, it gives you a great convenience, you can conduct a piece of research, you can see your audiences and then based on the audiences that are matching your brand, or matching your mission, you can then go straight to activation, you can explore those audiences with a single click, and they will be forming your social media channels, and then you’re ready to serve them the content that you want to serve.
So in a way, we are really making life easier for marketers, but in the future, our product will also look at researchers such as myself, because this product can not only be used for digital activation, are our ideas for this product to also be used in, you know, social research for, for instances where you want to do something in an agile manner, where you need a quick survey design. Our algorithm will be able to generate a questionnaire for your given brief. Then instead of spending weeks polishing the questionnaire, you will maybe spend a day you know finalizing that draft that the AI has generated for you. So that’s that’s in the principle of vision.
Matthew Todd
I think that’s a really, really interesting vision. if you compare it to, you know, a lot of the different surveying tools and things that are out there at the moment is, say, a massive step four, but it seems that it is, as you say, kind of allowing people to get access to that kind of information gathering without needing that team of experts behind them so that brands that might not have otherwise had access to this type of technology can then start to better find and engage with their audiences as well not just when they get to that, you know, massive scale where they got the resources to do that.
Mihajlo Popesku
Our vision and our philosophy is to democratize this, this technology, we want to enable the best in class research and digital activation to small and medium enterprises. We are starting with freemium, small brands will be able to take advantage of this technology. So the first piece of technology that we build, is what we call precision activation engine.
So this is essentially helping us to bring insights to life, when I say bring insights to life, bring insights to action, or platform currently can provide segmentation. Then you can explore desired segments to either Facebook and Google, and then have your audiences form and ready for activation. So that’s one thing.
Then second to that we want to build an agile research engine, agile research engine is something that only resembles survey hosting platforms where everything is, you know, do it yourself kind of thing. We want to build two things together with a survey hosting platform. One is a question bank that is going to be populated by a crowd of experts. Those questions will essentially be property of experts.
Every time our algorithm or our recommendation engine uses your question or questionnaire, you will be remunerated. So in a way, your question or questionnaire is your assets and a source of passive income.
Then secondly, that we want to build an AI that will help users of our platform generate questionnaires in a minute. Bespoke questionnaires in a minute, We don’t think about having, let’s say, having some kind of templates that you can reuse, we believe that every customer has a unique problem. Then specific to the problem to the brand domain industry country, the algorithm should be able to suggest an optimal sequence of questions so that your entire problem domain is covered. And so that you have your research questions answered in a full way, in appropriate way.
So our next step is definitely building that agile insight engine. But before that, we want to validate precision activation engine we want to see how it performs fine tune that. If there is an appetite, then we are going to proceed to this agile research engine. These are the plans for the next couple of years.
Matthew Todd
I can definitely see how you really isn’t innovative technology in that space. When it comes to bringing more users onto your platform, then how you’re approaching your own growth and marketing as well.
Mihajlo Popesku
So we recently started our sales and marketing. But then, you know, speaking, within the theme, we realize that the product-led growth is probably the west the best way ahead. We want our products to market itself. We want to, you know, give everything that we build and everything that we’ve built so far on a freemium basis, so that users can, you know, have a go at it, see how it works, and if they see any value in it, so that will be most important.
Then later on, once we progress to, let’s say, a paid version, so far, engine, we’re maybe thinking about charging a cut of marketing spend, that our clients will be spending, you know, using Qudo, or maybe having some additional features that will that will be charged. In the short to medium term, we’re, as I said, we’re thinking about the freemium model and letting the product market itself.
Matthew Todd
So you’re essentially, you know, being able to use those very, very large clients that you’ve got in financial services and other areas to allow you to build prove out and monetize what you’re doing so far. But then leverage what you’ve proven there to, to build this bigger scale product coming out of that development. Is that right?
Mihajlo Popesku
Yes, that’s right. So yeah, one side of our business is SaaS. But we also have a services part of our of our business where if a client wants something bespoke and then we can do something tailored for them, but if we’re talking about the SaaS business, yes, definitely looking at freemium and product like growth, and then building on that.
Matthew Todd
Thank you for for sharing that. It’s been really really interesting talking this morning. I know we had a couple of internet connection issues, but hopefully we weren’t too bad, but it’s been interesting finding out about the kind of marketing science approach, some of the limitations and the way you’re looking to itinnovate and push boundaries, as well as some of those trust issues and issues with first party data, as well, I think it’s really, really interesting.
Just to wrap things up, for our audience of, you know, entrepreneurs, startups, scale up founders themselves, is there any advice you would give them either in terms of building a business or in terms of their own marketing?
Mihajlo Popesku
You have to stay vigilant and flexible. Don’t hold on to your roadmap like a Bible. Be ready to change direction.
Mihajlo Popesku
Don’t keep ideas close to your chest. Share your ideas with people early on, let them you know, critique your ideas, let them question your ideas, find the good sparring partners, that’s one thing. Second thing, keep your mind open, and be ready to change your direction, you know, things are shifting fast, especially in technology markets are shifting fast, you have so many disruptions happening not not necessarily in technology, but also in wider like socio economic and political space.
You have to stay vigilant and flexible. So, you know, don’t hold on to your roadmap, like, like a Bible, like a gospel, be ready to change direction if you see that something, if you see compelling reasons to do so. Share your ideas, you know, let people you know, compliment you let people you know, give you their peace of mind. But at the end of the day, you know, if you’re an entrepreneur, you have to make calls by yourself. That would that would be my advice.
Matthew Todd
Fantastic. Thank you. I think that’s really good advice that anyone should pay attention to. Thank you again for sharing that journey so far with us. I definitely look forward to seeing the consumer version of the platform when you’re opening up that freemium model. I’ll certainly be looking out for that. I wish you the best of luck in that growth journey.
Mihajlo Popesku
Thank you so much, Matt.
Matthew Todd
Thank you for joining me on this episode of Inside the scaler. Remember for the show notes and in depth resources from today’s guest. You can find these on the website insidethescaleup.com. You can also leave feedback on today’s episode, as well as suggest guests and companies you’d like to hear from. Thank you for listening.