We recently connected with Peter Barcak, Founder & CEO of credolab, which claims to be today’s largest developer of bank-grade digital scorecards and data enrichment solutions based on behavioral metadata.
The company leverages privacy-consented and anonymous smartphone and online metadata to provide lenders, risk officers, and marketers with highly-predictive behavioral data.
With over 240 million digital footprints and 30 million scored individuals to date, credolab‘s toolkit is already being used by over 200 corporate clients across nine verticals. In this interview, Peter Barcak talks about the technology behind credolab, the advantages of using behavioural data, and the future of lending.
Our conversation with Peter Barcak is shared below.
Crowdfund Insider: Could you describe in detail how credolab’s technology works and its specific impact on risk management, fraud prevention, and marketing insights?
Peter Barcak: With the help of machine-learning algorithms, we analyze millions of data points from smartphones and online web behavioural metadata to assist financial institutions such as banks, fintechs, and BNPL companies in making more informed decisions regarding risk, marketing, and fraud. Initially, our vision was to redefine risk management by harnessing alternative data sources to enhance the risk underwriting process.
Credolab’s risk scores are designed to help lenders evaluate the creditworthiness of potential borrowers, thereby reducing delinquency and portfolio risk. Fraud insights aid businesses in the onboarding process or loan origination. Additionally, we have incorporated marketing insights to assist in targeting their campaigns more effectively, thereby decreasing the cost of customer acquisition and providing a comprehensive customer persona.
Crowdfund Insider: What differentiates credolab’s approach from traditional credit bureaus?
Peter Barcak: To put it simply, every individual possesses unique characteristics. However, banks and other financial companies often adopt a “one size fits all” approach. We believe that this generic approach fails to account for the diversity among individuals, as each person can interpret a generic message differently. Traditionally, financial institutions rely on credit bureaus, credit scores, or credit reports to refine their customer assessment process. Our technology utilizes anonymised data from device features, typing patterns, and session duration, among other factors, to provide insights for credit assessments.
We analyse behavioural data and digital footprints to identify patterns that describe how customers behave in real life, not only on paper. Based on this analysis, we derive a score similar to a FICO score, which lenders can incorporate into their underwriting systems. This ultimately leads to a model with greater predictive power, which is the value we bring. For example, instead of experiencing a 30% loss rate with the initial batches, they might observe a drop to 20%, representing a significant reduction in potential losses before fully understanding how to navigate the new market.
Crowdfund Insider: Does credolab’s data provide institutions with the capacity to experiment with market entry strategies and focus on successful approaches while disregarding ineffective ones?
Peter Barcak: Indeed, in markets where credolab is already present, companies can easily tap into our platform and instantly receive scores or insights through a single API call. This integration enhances their existing models and delivers immediate results. As for newcomers to the market, they typically integrate our SDK library into their apps or websites, allowing us to access the digital footprint of every individual they interact with. After a few months, once the portfolio has matured, they can identify reliable payers and delinquent clients within different client groups. It is worth emphasising that we already have a strong presence not only in the US and Europe but also in Latin America, Asia, the Middle East, and Africa.
Crowdfund Insider: How did the idea for credolab come about, and what was the journey of expanding the company after finding initial success?
Peter Barcak: Approximately ten years ago, I came across an article detailing how the CIA effectively used cell phone data to track down terrorists in Pakistan after the September 11 attacks. At that time, flip phones were prevalent, and smartphones were not widespread. The CIA obtained data from mobile phone operators and analysed patterns in the cell phone data of suspicious individuals, such as the number of calls made or the number of SMS messages sent. They discovered that these individuals were not engaging in lengthy conversations but were instead exchanging missed calls, among other notable characteristics.
The breakthrough came when they analysed all the cell phones in Pakistan and identified phones with similar patterns, which happened to belong to the terrorists they were tracking. This story resonated with me, and fast forward two years, and after spending nearly two decades in various risk management roles, I found myself in Singapore, where I decided to establish credolab. At that time, I was unsure if cell phone data could be used to predict credit outcomes or if accessing such data was feasible. However, it has proven to be a successful venture thus far.
As for the expansion of credolab, after developing our first scorecard in 2016, we began exploring various use cases and tailoring solutions to meet unique customer requests. Having invested in marketing efforts, we soon started receiving calls and emails from around the world, including Asia, Mexico, the US, the UK, Europe, and Africa. Although we did not plan for expansion at that stage, we managed to sign deals remotely from Singapore. This realisation prompted us to consider a different approach. Rather than building a vast sales network, we established one or two centres in strategic locations such as America, Asia, or Europe, allowing us to cover entire continents due to the scalability of our solution.
Our global expansion was further accelerated when the pandemic hit in 2020. Surprisingly, we managed to bring in more new clients than in previous years, even without the ability to travel. Based on this experience, we successfully raised funding for our Series A round and established small teams in key countries like Mexico, the US, and Brazil. The journey of credolab has seen both extreme lows and remarkable moments. For instance, in June 2020, our revenue dropped by 90%, but we raised our Series A round just a few weeks later. It is a testament to the fact that highs and lows can occur very close to each other in the startup journey.
Crowdfund Insider: Financial inclusion is a critical issue globally, with a significant portion of the population still excluded from formal financial services. How can lenders leverage your technology to expand their reach and promote financial inclusion?
Peter Barcak: Over the past decade, fintech innovation has enabled 1.2 billion unbanked people to gain access to financial services, according to the World Bank. However, there are still around 3.5 billion people worldwide excluded from the formal financial sector, most of whom are unbanked and rely on expensive informal credit sources.
Traditional and digital lenders can definitely extend their reach to underserved populations by leveraging alternative data sources like those that credolab provide. Traditional lenders can embrace digitalisation to simplify application processes, accelerate approval and disbursement times, and offer flexible repayment options. The key is to create a seamless and convenient customer experience that meets evolving consumer expectations. Strategic partnerships and collaborations with fintech companies can also help traditional lenders tap into their expertise in user-centric design and agile operations.
How this can happen is vividly illustrated by the success of super apps and fintech companies targeting the underbanked population in Asia. Superapps like Grab, WeChat, Line, and Alipay have gained tremendous popularity due to their ability to provide one-stop-shop solutions for everyday needs. In Asian cultures, efficiency and speed are highly valued. Superapps offer loyalty programs and discounts and localise their services to cater to local preferences. For instance, Grab in Thailand offers different promotions than in Indonesia.
Several countries, including Brazil, Indonesia, Mexico, and the Philippines, follow in Asia’s footsteps regarding their banked/unbanked populations, youth population, GDP per capita, and smartphone and internet penetration. Buy Now Pay Later (BNPL) services have gained popularity in these regions, addressing consumers’ lack of credit card ownership. By adopting innovative solutions and focusing on customer needs, traditional lenders can promote financial inclusion and cater to the diverse requirements of underserved populations.
Crowdfund Insider: Apart from BNPL, what other categories of lenders do you think should pay attention to behavioural data for better decision-making?
Peter Barcak: Early Wage Access (EWA) providers should be mindful of consumer debt issues and ensure responsible lending practices. Behavioural data can offer valuable insights for these lenders to assess creditworthiness and make informed decisions, thereby contributing to a more inclusive and responsible lending landscape. The EWA concept is simple and compelling. It gives employees access to money they have already earned, usually via third-party apps or payroll providers, ahead of their regular payday. However, while both these products are helping millions of people meet their short-term financial needs, they also raise issues about consumer debt.
EWA providers are convincing themselves that the excellent earnings data they get directly from employers make up for “thin files”, where there is very limited understanding about the person accessing the money. Just because you can see employees’ accrued income does not necessarily make them suitable subjects for early payments. And what if the employer goes belly up between any early payment and the actual pay run? That removes the employee’s ability to repay the advance. If the EWA industry is not careful, it might lead to another consumer debt problem like we see in BNPL sector now.
Crypto lending is another vertical to look after. The advances in risk mitigation achieved using analytics by conventional banks and credit card issuers have not fully filtered through here. It’s a puzzle, as crypto lenders have an even greater incentive to address risk, as blockchain-based currencies are well-known for their volatile nature. Things are starting to change. We are working with crypto lenders who are keen to learn lessons about risk mitigation using behavioural analytics. But there is still a wide gulf.
Crowdfund Insider: In your opinion, what emerging technologies have the potential to transform the customer experience within the lending industry?
Peter Barcak: Artificial intelligence (AI) and machine learning (ML) have certainly turned things upside down. These are undoubtedly handy technologies, but let’s not forget the creativity of the human mind. We can and should mix the achievements of different sciences and technologies. For example, credolab’s new personality-based targeting solution harnesses ML algorithms combined with the OCEAN model’s power, incorporating behavioral psychology to create highly personalised and engaging customer journeys.
This approach blends smartphone metadata collection, machine-learning algorithms, and insights from behavioral psychology. This allows us to gain a deep understanding of the personality traits of top-performing potential customers.
It’s crucial to note that this technology is not about browsing history or keyword searches. All insights are gained without any personal data or identity markers. Anonymised smartphone behavioural patterns are being collected, such as charging habits or screen brightness, which provide unique insights into the personality type of the user. The credolab team has already analysed 1 million data sets and built a robust framework that is currently deployed in production.
This technology has already caught the attention of major industry players, including a large online booking platform in Latin America with millions of app users and a digital bank in Southeast Asia with 3.5 million users. Some proven results include a 32% reduction in Cost of Acquisition (CAC), a 22.9% increase in Return-On-Marketing Investment (ROMI), a 58% improvement in Intent Detection, a 31.3% increase in Return-On-Ad-spend (ROAS), and more.
