Tekios spoke with Ivan Caballero, CEO and founder of Citibeats, Ibero-America’s govtech revelation.
When the Spanair airline went bankrupt, Ivan Caballero also crashed.
Pained and resigned, he put the magic coin he had created to make the world a better place in a drawer. The Social Coin was virtual “dough” with which Caballero convinced thousands of people in 60 countries to do a good deed and receive another in return, thus forming a chain of favors.
But that dream lasted two years before ending quickly. Caballero could have taken some time to rest and slowly decide what to do next. Instead, he felt the itch of urgency to start over.
After three startups that ended like Spanair, he has no doubt that the leitmotif of the new venture, like that of The Social Coin, should be “I contribute.”
This is how Citibeats was born. Today, it’s a govtech platform with promising prospects.
Gone are the days when the startup had to play a seducer to convince potential investors. It recently received an investment of more than $2.3 million USD from IDB Lab (the innovation laboratory of the Inter-American Development Bank), CAF (the Development Bank of Latin America), Estinvest, Previsora General Mutualidad de Previsión Social, and Everis Fitalent.
Citibeats also partnered with the World Health Organization (WHO) to analyze data and content related to COVID-19. Its client portfolio includes organizations such as Twitter, Facebook, the United Nations Development Programme, the Inter-American Development Bank, the International Telecommunication Union, NTT Data, and the United Nations Educational, Scientific and Cultural Organization (UNESCO). All are eager for Citibeats’ data and advice, and the virtuous use it makes of artificial intelligence.
Citibeats has its eyes on Latin America, where it plans to open its first regional office.
Q: In 2016, you created the Social Coin. That year, many crucial things happened in Barcelona and Spain. There was a social crisis, there were issues related to mortgages, there were people protesting in the streets because they did not have money to pay the rent or buy food. All in all, the social, political, and economic model was in a crisis. Something “systemic” was happening that perhaps ended up catching your attention, becoming your call to action.
A: Yes, there were many problems. We were experiencing an economic crisis and issues with mortgages. The evictions, particularly here in Barcelona, provoked a lot of social unrest. However, the call to action for me was that I had just gone bankrupt. Citibeats is my fifth startup. The previous one (the Social Coin) went bankrupt because we had a large client—a Spanish airline—that went bankrupt overnight and dragged us down with it.
Q: What did the first three startups you created, before Social Coin and Citibeats, lack? What component did they miss?
A: Well, I’m a pretty intense person in everything I do, but what was missing was a purpose, a clear goal. I am very good at conceptualizing technological solutions, but these were not at the service of society. On the other hand, after the Social Coin experience, I decided that I wouldn’t do anything again that did not have a social meaning and purpose.
And at the same time (talking about new goals), well, I had found out that I was going to be a father...I said to myself: “If I start a new project, it has to make a lot more sense than what I have done so far”.
What is the goal of human beings on Earth? Contributing to society.
Q: But the Social Coin didn’t work. And again, it was necessary to rethink and redefine the purpose.
A: In fact, we started thinking that the Social Coin could be the starting point for a sustainable model that might work in large corporations through corporate social responsibility. But to do something like that, either you do it with an altruistic approach or it doesn’t work. This is how Citibeats was born, while the Social Coin was more like a tool to generate corporate culture.
At Citibeats, we help understand major social trends and give a voice to the maximum number of people throughout society. In other words, we create models that are as broad as possible so that the information processes used by large multilateral organizations—which are the ones channeling money flows to society—are much more efficient. Efficiency is measured not only from a time point of view, or considering the quantity and the quality of the answers collected, but also by incorporating many more voices (i.e. opinions) than traditional methods.
Q: At that moment, I guess that you had to face a new challenge: how to connect with people who make decisions on a large scale—municipal, regional, and global. Because it’s one thing to motivate and move those who are peers, right? But it is another thing completely to do the same with a sector of society with which most people do not relate. They make decisions, often from a kind of dome—remote and untouchable. How were Citibeats’ first approaches with that environment?
A: The first contacts were mainly with small cities or national governments. Our rationale was that those potentially willing to pay for our solution were essentially those who govern—those who make decisions for the people. What we rapidly understood, though, was that governments (at the national and city level) follow very tight agendas because here in Spain they have a four-year term to govern. So it is very difficult for entities like national or local governments to risk using such a ground-breaking and disruptive technology. We understood that it was not going to be easy at all.
Q: What did you do? What was your plan B?
A: We started collaborating with our partners. Everis and Telefónica were our investors, so we thought they could help us resell our solution. They have thousands of salespeople on the street who know all these governments, and they could help us resell our technology.
Q: What was your first project? To whom was it sold?
A: I fondly remember one of the first projects with the government of Navarra in Northern Spain. The goal of the project was to detect hate speech towards vulnerable, low-income communities, migrants, and, in some cases, against women as well. So, what we did was to identify these types of narratives to generate corresponding counter-narratives, mitigate any possible incidents, and prevent other similar narratives that could generate violence in specific neighborhoods.
Q: You collected sensitive information so that the government of Navarra could target its policy having a more precise idea of what actions were needed and where.
A: Absolutely, and we managed to generate a fully automated decision process. With our technology, we were able to work with different languages—not only Spanish but also Basque and French. This allowed us to geolocate the narratives so that the Navarra government could automatically assign a certain budget to a specific neighborhood.
The Citibeats Team
Q: Once the project ended, did any multiplier effect occur? Did the word spread about the results that Citibeats achieved in Navarra?
A: Yes. Actually, we started a collaboration with the Government of Catalonia. We also began a dialogue with the Instituto de Empresa (IE), which has a team dedicated to govtech initiatives. Together, we were able to have a better understanding of multilateral organizations. We also started working with IDB (Inter-American Development Bank) and the UNDP (United Nations Development Program).
Q: What was the Spanish govtech ecosystem like at that time? Was there an actual ecosystem, or were there just a few initiatives?
A: In Spain, the govtech ecosystem is still in a very emerging phase. At least, today. It is immature, the hiring system is still very traditional—very little adapted to the govtech phenomenon. There are few startups calling themselves govtech as well. It’s all still incipient. We don’t even call ourselves “govtech” because sometimes the concept might scare investors and certain types of customers.
Q: Why are investors scared?
A: They are scared because everything connected to the public sector usually translates into low recurrence, low scalability, poor money, and long customer acquisition times. Those are the four main pain points of the public sector. Govtech is just a modern way of saying “I work with the public sector,” and that terrifies them.
What we are discovering with multilateral organizations is that it is not true that they pay as badly as national or local governments. Once you have managed to earn their trust, they are quite quick to react and are very loyal. So, if you happen to have a solution that fits their priorities, it works even better—much better than with private companies.
Q: The govtech ecosystem is just emerging, as you say. To that, it must be added that the world of technological entrepreneurship seems to be in a sort of race. It is as if everything is an ego-fueling process with the goal of achieving the greatest bridge round. Does that contrast with what is experienced in the govtech world?
A: That’s a very good question. I think it is just a matter of trends. In other words, we are a govtech company but also an artificial intelligence company. The main difference is that all fintech companies, for example, are on the rise. They are riding a wave with gigantic budgets.
In the govtech sector, budgets are smaller because there is much more risk aversion. Rather than betting on a startup that is cheaper, a public official prefers to spend ten times his budget for a Deloitte or an Accenture even if they do a bad job because no one is going to question the work they have done. And if something goes wrong, well… why take the risk?
Q: What are the main cultural barriers that you detect in the internal structure of these governments, especially the local ones when they want to implement this kind of solution? The feeling is that research & development and IT are sectors where the borders are not really well defined. In some Latin American municipalities, for example, sometimes you see a qualified professional who is asked to do basic stuff, like setting up a printer. What are the obstacles that you had to face more often? Corruption, risk aversion, lack of specialized human capital?
A: We have never faced corruption. Really, never. On the other hand, aversion to risk is a great blocker. Why would someone want to work with a Spanish startup in Latin America? It is very difficult to enter that market unless you go hand in hand with a multilateral. Now it is much easier with CAF as one of Citibeats’ main investors and after strengthening our relationship with IDB (also an investor and a client).
Then there is the training aspect, but this does not apply just to municipal governments. It also affects national governments and large multilateral organizations. There are no big data or data science teams capable of using technology autonomously. We are moving towards being a SaaS (software-as-a-service), but it is not easy because the market is not mature enough and these organizations do not have the internal resources yet.
And then, there is a third blocker represented by the hiring model. Municipalities always have budget limits. Here in Spain, we are talking about €15,000 (roughly $17,000 USD). Plus, they cannot hire you for two years in a row. It is just one shot; one year and that’s it. A project might easily cost from €40,000 (about $47,000 USD), up to €400,000 (about $470,000 USD). Clearly, with all these blockers, everything gets much more complicated.
What is the goal of human beings on Earth? Contributing to society.
Q: Here in Latin America something else happens: the authorities have a plan and they start it. Then, those who replace them in the government usually say: “Hey, this project is not mine. Out!!” Does this happen in Spain?
A: It happens in all places. In Spain, we have it every four years. And it is a problem because in those countries or regions where the governments are constantly changing, it is impossible to build anything that lasts even a bit, that you can make grow little by little. The general tendency is to behave in an opportunistic, selfish way.
Q: And it is even more complicated if the change is from one political party to another that has opposite views.
A: Of course, because in the end, what really matters for everyone is the credit you get. I mean, if I am responsible for a winning policy, then I will get all the credit and my reputation will increase. If I am not reelected, I don’t want someone else to benefit from my political decisions and actions. It is a four-year term, so we try to limit the number of challenges we face and win them all within those four years.
Q: With CAF’s first govtech investment in Latin America, Citibeats must start looking for opportunities in South America. What do you know about the culture? Do you know what challenges you will need to face? What do you think the future holds for you?
A: The region is teaching us a lot. It is a complex area because, yeah, we talk about Latin America, but there are many countries, languages, and cultures. There are also differences within the same country.
Q: Despite the differences among so many countries, have you detected common pain points in the region? Maybe some emergencies?
A: Well, what we are mainly detecting, right now, are issues closely related to inequality and violence due to COVID-19. We are focusing a lot on analyzing the disinformation narratives because we realized that there are groups that do not have access to information or that are driven by trends or gossip (rumors). All of this makes them vulnerable communities.
Q: That means that Citibeats has already begun to confront the “plandemic” supporters.
A: Yes, actually, eight months ago we launched the first disinformation observatory around the pandemic. It is a public observatory that anyone can access to consult the information collected. We are in the process of expanding and escalating the project so that it can be used to monitor other pandemics or other diseases, such as Ebola.
Q: And all this is possible thanks to a mix of technology and data. What are the sources that Citibeats uses? Where do you collect your data?
A: Citibeats is a social intelligence platform that helps process large volumes of text-based data—any type of data in any language—in order to extract insights in real time reflecting changing social trends. Citibeats can work with any data source. Social networks, forums, websites, blogs, news, comments, transcripts of calls to the citizen service telephone line. We gather all these data from different sources and then anonymize them. This is very important to us because it is one of the key points that make our artificial intelligence technology ethical and responsible. The data sources are unbiased or have the least possible biases; data is anonymous, and technology is used to improve people’s lives.
Q: The topic of ethical artificial intelligence has vital importance when it comes to govtech matters related to security, public safety, excesses, protests, etc. These are all examples of areas where Latin America’s central and local governments and even municipalities need to act. For six months, there were daily “social explosions” in Chile, and the same happened in Colombia. Is security an area where you could get involved, or do you think your technology doesn’t apply to this area?
A: Yes, we could, and we will absolutely get involved in security matters. In fact, one of our clients asked us the following question: when is the next social outbreak going to happen again in Chile?
Based on the original Tekios article published in Spanish.