How can you be sure someone will repay their credit? That’s the million-dollar question for any organisation issuing loans. To evaluate creditworthiness and reduce the risk of default, decisioning models had to be created.

Statistical models have been in use since the 1950s. In 1941, David Durand, a management professor at MIT suggested using statistics to help with credit decisions, particularly for private loans. In 1956, the engineer William Fair and the mathematician Earl Isaac worked on these statistical models for granting credit. They founded a business, Fair, Isaac, and Company (FICO) – where the famous FICO score in the US comes from – and produced their first scoring model in 1958 for the American Investment and Finance Company. 

These statistical scoring models were deemed to be a lot more effective at separating good from bad risk profiles than a model that was based only on the judgment of employees working in credit establishments. This has helped maximise the share of ‘good loans’. The models have become more sophisticated over the years, but the principle remains broadly the same. They continue to be based on a combination of demographic, socio-professional, and financial data.

After more than 60 years of loyal service, scoring models deserve to be reinvented. Particularly with the emergence of more complete financial data on the market – Open Banking data. With this as a base, new scoring models are possible: API-based behavioural Credit Scoring. This allows a better credit risk management but it also opens access to credit to people who have been excluded from it up until now. This is what we are working on at Algoan. Youness Bounif, our Chief Product Officer reveals how behavioural credit scoring using APIs is reshaping the landscape of the credit sector. 

The power of Open Banking data, accessible via API

There has been a new type of financial data available for a few years: Open Banking data. This new type of data is whether regulation driven or business driven.-

  • In Europe it was the Payment Services Directive 2 that pushed banks into developing APIs. These enabled third-party services like fintech to leverage customer banking data (with their consent). 
  • There was a different approach in the US; the data revolution was led by the players themselves who spotted the business opportunities presented by Open Banking. 

Open Banking is incredibly powerful whichever approach is taken and has enabled the creation of data models that are internationally scalable. Why? 


  • Data universality
Before Open Banking With Open Banking
Scoring models are based on very disparate data that is not just financial. There are large variations between countries. It is a mix of demographic, socio-professional, and financial data. Data is the same in every country with Open Banking in force. It is the banking data of individuals: accounts and transactions (description, amount, date, etc.). It is found everywhere.


  • Processing universality
Before Open Banking With Open Banking
Processing differs according to various elements (single versus in a couple, tenant versus homeowner) depending on the country. Every lender has their own individual credit scoring. Open Banking means the assessment of credit applicants is carried out in the same way, whatever their country of origin. The assessment is done by studying their recent, up-to-date financial behaviour.


  • The end of silos between countries
Before Open Banking With Open Banking
Some countries have credit bureaus that give access to an individual’s credit history and any related repayments. In the UK and the US for example, this information is made available when an individual applies for a loan. France doesn’t have a credit bureau, and thus no credit score or repayment history is available, just files showing those who are banned from access to some banking services. If a lender wants to operate in several countries it needs to tailor its model to accommodate local scoring practices – a mammoth task. Credit scoring operates in the same way in every country. Replicating models in different countries is straightforward. It is possible to add a layer of extra personalisation to a common base and so accommodate the local context and further refine the scoring model. Partners like Algoan handle all the details. API credit scoring facilitates a global presence without the customer needing to take on any integration work.

 ”At Algoan, our product uses Open Banking data exclusively. We tailor our algorithms on a country-by-country basis so that scoring is optimised and tailored to each country. It is completely transparent for our customers as Algoan handles all the complexity.”

The benefits of a turnkey solution (APIs)

So, credit scoring by API uses powerful data that give a true picture of the financial behaviour of borrowers. The share of ‘good loans’ is optimised while also allowing people who have previously been excluded from credit (e.g. the self-employed) to access it. 

API credit scoring disrupts the market by delivering three significant advantages to lenders.

  • Eliminate barriers to entry for lenders

Why were there very few organisations offering loans besides banks until recently? The reason is simple: the difficulty of entering the market. For example, in France, without credit bureaus or credit scores, this is used to mean the development of individual scoring models based on socio-demographic data. On the one hand, this is difficult to build and on the other hand, it is not very exact. Even in countries that have credit bureaus, entering the market is hugely expensive (€50 000–€500 000  depending on the country). This represents a huge barrier to entering the French market or any other market.

With API-based behavioural credit scoring, the borrowers’ solvency and creditworthiness are assessed using financial data that is reliable and universal. Barriers to entry for new entrants to the credit industry are disappearing, and that naturally opens the market to competition. It is good news for consumers who have more choices.

  • Make credit scoring convenient

Developing an individual scoring model used to be a major blocker for lenders. The skills are complicated to acquire, yet it is necessary to ensure that default rates are not too high. In some countries, like the United States, credit scoring is already convenient with the FICO score. Only, it doesn’t cover the whole population. People who have never applied for credit are not included, or anyone from abroad. They are outside the ecosystem and so excluded from credit.

Almost everyone has a bank account from which they make transactions. With Open Banking data, credit scores are entirely based on a person’s financial behaviour. The API-based credit scoring developed by Algoan is more precise than other scoring models and doesn’t require any further effort on the part of the customer.

  • Facilitate international development

Until now, replicating a scoring model in a different country was heavy going and (very) expensive. Models had to be built from scratch, and the cost of implementation was exorbitant. Just imagine the entry ticket the credit bureau requires was multiplied by the number of countries in which the organisation wanted to develop. With Algoan’s API an organisation can set up in a new country almost instantly.

“Going from country A to country B is easy and doesn’t involve extra costs with API-based behavioural credit scoring.”

The strength of the Freemium business model and API calls

Want to test a solution before buying? It’s impossible in the credit world. At least it was before Algoan. 

Until now, to start using a credit scoring model you had to discuss it with the company, negotiate (lengthy) and pay (expensive). We have opted for a Freemium business model, ie. a portion of our offering is free. 

“Startups, fintech, credit and BNPL companies can implement our scoring solutions in an autonomous, self-service way. Our free service includes a limited and specific number of API calls. If some have limited credit scoring use they can stay in free mode. For more functionalities and services there are our paying options based on usage.”

Why have we chosen a Freemium model? It allows companies to try out our solution without any price barrier.

“It is completely normal for customers to want to test the credit scoring product before committing. It is often central to them. We’re sufficiently confident in our product to offer it for free. Once customers are convinced by their first tests they can retain free access if their use is limited or move to our paying products if they have more volume or they want more functionalities and support.”

We have also adopted a pay-as-you-go model where you pay for what you use.

Embedded finance is a strong trend in financial services. This offers off-the-shelf solutions while integrating technology modules built by other businesses. This is what we are doing at Algoan with our credit scoring API, particularly with fintech. They often position themselves as international from the outset. We support them with their international development by offering them a credit scoring model that they can use to conquer new markets almost instantly. API Credit Scoring works better and works everywhere.