In recent years, lenders have discovered the benefits of using alternative credit score providers to improve the predictive power of their risk models. Using scores based on these advanced data sources makes it possible to provide a new depth of understanding a person’s creditworthiness. Nevertheless, it is important to note that not all alternative data is the same, and not all providers offer equivalent data.  

‍

Credolab’s solutions are rendered as complementary to alternative data scoring solutions that may use other forms of alternative data, such as telco, psychometric data, bank account-related data, or utility bill data. In addition to affordability and propensity for repayment, these data sources enable insight into creditworthiness from different perspectives. Furthermore, with the proliferation of these providers, it is possible to leverage multiple alternative data sources at a time, improving the predictive power of their risk models.

‍

About telco data

Today, with smartphones' rapid expansion, telecommunication companies are collecting large amounts of data. This includes mobile phone usage, call and SMS records, network status, server logs including SIM card swap, billing and top-ups, and roaming data.  This data holds useful customer insights that, when analysed accordingly, can help financial institutions reduce credit risk, prevent fraud, and increase approval rates.

‍

Telco data, for instance, can be extremely useful for preventing fraud. Many telco data providers offer fraud detection flags by comparing geolocation data with the information that the applicant declared during the onboarding process. For example, a mismatch between the home and work address with the geolocation address obtained from the user's cell phone tower could be a warning for suspicious behaviour.

‍

How credolab is the best ally to telco data

Credolab offers a new depth of information to most alternative data providers since it is based on behavioural data. In particular, the data is collected in the form of anonymous metadata (without any PII ever leaving the device), in a JSON format, and processed to assess the customer’s creditworthiness of any applicant and detect fraud. This form of data offers new knowledge about the subscriber, which is their willingness to repay, besides the affordability and the propensity to repay.

‍

Credolab is a great complementary solution to alternative telco data since behavioural data is layered on and has a very low correlation. Moreover, being a leader in the field, credolab analyses nearly 70,000 privacy-consented and permissioned data points across various data categories, whilst telco data may reach just about 1,000. Let’s see in more detail how metadata can be used by credolab to evaluate risk and fraud and also to enrich marketing segmentations:

‍

• On mobile, credolab collects permissioned metadata across Device Information, Registered Accounts,  Contacts, Calendars, External storage (media), Internal storage (media), Application (Android only), and User Interface interactions. This also includes the total time spent applying for a loan or credit card, the time spent in the same position, Latency, and Keystroke patterns.

• On the web, credolab collects Device and Browser Information, Language and Operating System metadata, and User Interface interactions, including the total time spent applying for a loan or a credit card, the time spent in the same position, Latency, and Keystroke patterns.

‍

All of the above are leveraged for a thorough analysis of each customer to determine their solvency and creditworthiness in a more granular and accurate way.

‍

What is the difference between credolab and other providers?

With Credolab, mobile device and web behavioural metadata are used more elaborately and extensively than most alternative data vendors. Credolab uses only privacy-consented, permissioned, depersonalised and anonymised metadata, while most providers access and extract personal and sensitive data (PII).  At the same time, credolab possesses the know-how to behavioural feature engineering, which is at the core of its AI-driven algorithms.

‍

Credolab can also understand user behaviour and interest using insights obtained from mobile phones and web metadata. For instance, through the analysis of mobile app ownership, credolab could discover that the prospect is an online gamer, which could indicate a lower propensity to repay, or, on the contrary, they could monetise consumer purchase intention by monetising it by offering more tailored products or services.

‍

Credolab’s solution can also provide insights useful for UX teams related to the depth of a client’s tech savviness that other alternative vendors can’t provide. For instance, by detecting the number of times the customer changes a certain field or uses copy/paste/delete in filling out an application form, the lender can detect onboarding pain points and eliminate friction, improving customer experience.

‍

How does credolab add economic value to the existing prediction models?

In short, it is evident that every company should be using as many data sources as possible as long as

‍

1) They add marginal value to the predictive power of their model, and

2) The price to ingest each additional data source makes sense from a unit economics point of view.

‍

Through unit economics, it is possible to break down a lending business into smaller components to see how it performs as a whole. This includes the average cost of servicing (administrative costs) and, the average cost of financing, how much the lender pays for those lent funds. So, as long as companies like credolab can keep the approval rate high, the probability of default low, and they increment marginal value, there is no reason why a lender should not use more alternative data.  

‍

A McKinsey report published in June 2017, Risk analytics enters its prime, explains that for every 1% uplift of predictive power, the lender saves 1% in the cost of risk. At credolab, we know that prediction can be improved by anticipating behaviours and focusing on deeper level insights besides traditional patterns historically utilised by credit bureaus. A behavioural data-based approach helps credolab enhance predictive models by providing new insights that help lenders attract new clients and cut acquisition costs whilst in compliance with user privacy regulations. At the same time, prediction can be further improved by layering on other relevant alternative sources, such as telco data.

‍

Interested in learning how our products can help you? Request a free demo, or drop us your questions here.

‍

‍

Access data insights solutions that deliver growth - Fraud detection | Credit scoring | Marketing segmentation. Helps you say "YES" more confidently to more customers!

‍

Learn more about credolab's products and possibilities with our features through our Blog section, and feel free to share our content with your team!

‍

‍Follow us on social or get in touch today: Book a meeting | Blog | LinkedIn  | Twitter |Contact Us