Categorization of data: The key to hyper-personalized banking

The information contained in transactions has become a true goldmine, capable of revealing consumer preferences and business opportunities.

Table of Contents

However, to harness its full potential, it’s necessary to go beyond simple data collection and delve into categorization and enrichment. Coinscrap Finance has been helping the financial industry hyper-personalize their customer relationships for years thanks to their own AI engine, COCO

Juan Carlos López

Co-founder & CPO of Coinscrap Finance

What is financial data categorization?

Financial data categorization is a 100% digital and automatic process that allows financial institutions to classify each user’s bank transaction into a defined hierarchical structure: category and subcategory.

This means greater expense understanding for individuals, identifying the merchants where they are buying and displaying the type of business establishment: leisure, health and sports, insurance, car and transportation, etc. 

Subcategories add an additional layer of information, indicating if it’s a concert, a gym, home insurance, or a gas station. This enables people to make more informed financial decisions in their daily lives.

How is categorization used in banking?

Grouping income and expenses into categories and subcategories makes it easier for users to analyze their household economy. Banks can thus provide a structured view of the merchants where the customer spends most money.

In this way, individuals have all the necessary information to manage their finances optimally. Additionally, by providing banks and insurers with a detailed view of customer behavior, we contribute to the generation of personalized recommendations and the improvement of business strategies.

The secret to an enhanced financial experience: Types and specialties

Coinscrap Finance goes beyond simple data categorization. Through our solution, we go one step further, adding valuable information to each transaction: type and specialty.

The process of enriching transactional data takes just a few milliseconds. With it, banks and users obtain an extreme level of detail: merchant name (or brand), logo and URL, economic activity, address, geolocation, payment method used, type of establishment, specialty, if it is a recurring payment, and the carbon footprint associated with the purchase.

As a recent study by Pronix Inc. mentions, the secret to increase retention rates and reduce acquisition costs lies in using data to offer a hyper-personalized experience to each customer. Now banks can recognize relevant events in their users’ lives due to technology.

Additionally, by displaying all the additional information that types and specialties offer, customers can navigate through their spending history, identify consumption patterns, and even detect possible fraudulent purchases.


Banks also benefit, as chargebacks are reduced by up to 25% thanks to greater customer understanding of transactions. Recently, our UK neighbors made public a striking fact: the total value of refunds and disputes over charges exceeds 10 billion pounds per year.

New call-to-action

Artificial Intelligence at the service of transactional categorization

Behind detailed categorization is a powerful artificial intelligence engine: COCO. As my colleague Óscar Barba explains, COCO has been developed in collaboration with the University of Vigo.

This system, a leader in banking data categorization in Spanish, with more than 90% accuracy, can process millions of transactions in the blink of an eye, ensuring an unparalleled user experience.

In this way, we can increase the financial health of the population and help banks create intuitive applications that enhance the user experience. At Coinscrap Finance, we believe that facilitating the understanding of personal finances empowers people and enables them to achieve their savings goals quickly and effortlessly.

Tangible benefits of categorization for financial institutions and users

Outline of the categories, subcategories, types and specialisations of transactional data categorization.

The competitive advantage of knowing the end of a subscription, the expiration of insurance, or the arrival of a new family member, revolutionizes the rules of the game. All this and more is contained in the customer’s transactional information.

Identification of subscriptions and recurring payments

Let me tell you something about these two aspects. By analyzing the name, amount, and date of the charge, our tool can identify these types of payments and return valuable information. Subscriptions are small expenses with significant weight in the family economy, so keeping them under control is crucial.

Whether identifying a video streaming service (Netflix, HBO, Disney), audio (Spotify, Amazon Music, Apple Music), press and magazines (Expansión, Marca, Telva), or education (Udemy, Domestika, Babbel), the ideal is to be able to see this type of expense in one place and in a 100% digital way.

In the case of recurring payments, such as VAT, insurance, or bills, it is very useful for the user to receive alerts in case their account might go into the red due to an imminent payment of this type. Institutions position themselves as allies by offering advice tailored to each person’s specific characteristics and delivering it at the right time.

Technical aspects: coverage, scalability, and integration

As I mentioned, the fact that our AI engine has been training in Spanish for years implies efficiency rates very close to 100%. This allows Spanish and Latin American banks, neobanks, and Fintechs to take advantage of our knowledge simply by connecting to an API.

One of our key markets is Mexico, from where we continue to grow towards countries such as Argentina, Colombia, or Chile. As a Galician startup, we also aim to pay special attention to the Portuguese market, our neighbors and strategic allies.

This global coverage, combined with the scalability of our tools, allows us to meet the needs of a wide customer base, regardless of location or organization size. 

Our implementation is characterized by its simplicity and flexibility. This platform integrates, as mentioned above, through an API, allowing companies to incorporate this technology into their own applications swiftly and without complications.

The future of banking data categorization with artificial intelligence

Transactional analysis has become a fundamental piece for the transformation of the financial industry. Many entities are discovering that it is necessary to jump on the innovation bandwagon not to be left behind.

I hope that knowing about our categorization, enrichment, and optimization platform for transactional data has helped you understand the real importance of customer satisfaction.

According to a Movizzon study, banks lose 20% of their customers due to poor user experiences. Caring for contact with digital platforms is crucial, as new generations practically only interact there with their bank or insurer.

With the evolution of technology and artificial intelligence, entities have allies like Coinscrap Finance to keep their customers satisfied throughout their lives, helping them improve and protecting their financial health. That is the secret to continued trust in a specific brand year after year.

New call-to-action

About the Autor

Juan Carlos López Díaz is Chief Product Officer and co-founder of Coinscrap Finance. In 2016, together with David Conde and Óscar Barba, he created Txstockdata and later, Coinscrap App. After the enormous success obtained, the business pivoted towards B2B with the help of EVO Banco and its “Smart Piggy Bank”. As a developer, he has more than 8 years of experience leading large projects.

Together with his team, he is able to create the best tools for the world of finance. Throughout these years, he has carried out important technological integrations for large companies in the sector: Evo Banco, Santander, Caser, Mapfre, Aon or Bankia. Juan Carlos is an Electrical Engineer from the Central University of Venezuela and iOS App Development from the U.N.E.D and the U.C.A.M.

Search
Subscribe to our newsletter

Do you like the content? Subscribe and receive our biweekly newsletter directly in your inbox.