We use cookies to give you the best possible experience on our site. By continuing to use the site you agree to our use of cookies.  

About

FriendlyScore is a SaaS B2B solution that analyses social and online data to assess borrower risk and default probability. This helps lenders increase approval and conversion rates, and offers borrowers better access to financing.

Our values

values icon first

Financial inclusion for young, underbanked, and internationally mobile populations

values icon second

Objective credit decisioning based on algorithm scores, not human biases

values icon third

Better conversion and approval rates for lenders through supplementary social data

values icon fourth

Consumer data protection from user-permissioned data access; no personalized storage or reselling

Thought pieces

Financial Inclusion
Why social network data?
Scoring methodology
Consumer privacy
Our Purpose and Values
Five applications for alternative lending data
Want to find out more? Read our Medium profile

Leadership Team

Loubna Bazine photo

Loubna BazineCEO

Loubna is an experienced finance professional, with many years of experience in commodities derivatives in major institutional banks including Bank of America Merrill Lynch, Standard Chartered Bank and Mitsubishi Bank. She also worked as a Senior Fundraiser in the Fintech sector. As our CEO, she brings a wealth of knowledge and valuable expertise which she acquired over the course of a successful career. Loubna holds a master degree in Mathematics from Columbia University in New York.

linkedin icon
Emilian Siemsia photo

Emilian SiemsiaCTO

Emilian is a web developer specialising in PHP, Python and JAVA. He has a special interest in artificial intelligence, particularly semantic analysis. Pior to starting FriendlyScore he ran a software development house. Emilian was educated at Wroclaw University of Technology.

linkedin icon

Knowledge base

1. What is social credit scoring?

Social credit scoring is an alternative approach to credit scoring that is used to assess thinner-file borrowers such as students, foreign nationals, and underbanked populations. While traditional credit scores work well, they exclude vast portions of the creditworthy population who lack historical data. Social media and online profiles are hubs of personalised, internationally standardised, verifiable data that can be easily packaged for lenders.

2. What is FriendlyScore?

FriendlyScore is a SaaS B2B scoring solution that allows online lenders to perform user-permissioned checks on borrowers based on social media and online footprints, IP verification, and browsing history analysis. We gather data from LinkedIn, Twitter, Instagram, and Facebook, with the imminent additions of Paypal, and bank APIs. We also verify with select external databases.

3. What social networks are used in social scoring?

We currently use four social networks: Facebook, LinkedIn, Twitter, Google. We’re also improving our scorecard to allow users to select additional online profiles to use for scoring.

4. Is FriendlyScore consistent across all markets?

Our basic scorecard focuses on internationally standard online platforms, but we can easily add market-specific social scoring features.

5. In which lender markets has FriendlyScore been adopted?

UK, Germany, Holland, Poland, Romania, Brazil, South Africa, Nigeria, India, Philippines

6. Where and when is FriendlyScore most effective?

The most predictive scores come from countries with high social media penetration. Consumers in these markets tend to be active on multiple social platforms. The problem we are solving is largest in countries with limited credit bureau data, such as emerging markets. In such places vast portions of the population can only be assessed using alternative methods. So if such a country also has high internet penetration, online and social data is an effective way to assess borrowers.

7. What is required to calculate a FriendlyScore?

FriendlyScores are user-permissioned. Users must allow FriendlyScore (or the white label lender) to access their online profiles. Should photo identification also be required, users may need to independently submit an identification picture.

8. Do users see raw data or just their FriendlyScore?

Users only see their calculated FriendlyScore and a few interesting facts about their social media behaviours.

9. What are the Scorecard Variables?

FriendlyScore classifies the scorecard input variables and the continuously observed algorithm data into three categories:

1. Basic Personal Data – fundamental personal data like demographics, employment, education, and residential history that can intuitively be used for social credit scoring (some of it has been used in traditional models). However, FriendlyScore can leverage larger and more diverse population segments by using alternative data sources like social media and online footprints.

2. Social Network Data – data reveals the strength of the user’s social networks, connections with family, professional contacts and levels of engagement, which can be further enhanced by our premium endorsement feature.

3. Web activity - data shows time spent on web pages, geolocation, IP checks, financial product browsing, and fraud checks with external data sources.

10. How is the model backtested?

We learn and grow with every customer. The data we collect improves our general scorecard and helps us customise scorecards for clients focused on unique markets or products. We ask clients interested in scorecard customisation to share the performance data of every scored borrower, even in a simple format like 0/1/2 (0:default, 1:paid late, 2:paid on time).

Custom scorecards must have at least two indicators (good/bad), but we advise clients to establish more criteria; more performance indicators will strengthen score results. Custom scorecards activate after the initial 1000 borrowers are scored. Further customization is possible after every 3000 scored borrowers thereafter.

Our API connection allows for seamless, automatic retrieval of performance data. If an API connection is unfeasible, we request manual performance data sharing.

11. Can an offline lender use FriendlyScore?

We bring offline lenders online by offering a white label, customisable landing page that can perform credit scoring and manage user-friendly loan application. Once a user is scored, you receive an auto-generated email with their profile and basic loan data which can be used to processed offline.

If you engage borrowers in-branch (not online), we can serve them through POS device integration or text message invitation.

12. What does it cost?

FriendlyScore rates are tiered based on volume of people scored. Please see pricing

13. What initiated the creation of FriendlyScore?

Technology businesses are making financial services cheaper, faster and more accessible by democratizing supply as well as demand away from traditional banking oligopolies. Open marketplace platforms have lead a massive expansion in the way people can access financial services in many forms such as peer-2-peer consumer lending, enterprise crowdfunding and international cryptocurrency payment networks to name a few. Large banks, in some cases, have started to innovate and partner, in order to have good online products as well

Little progress however has been made in credit-decision capabilities, an essential next step towards greater financial inclusion. Traditional credit bureaus use old and conservative measures of affordability, such as number of years spent in the same residence or job. While these measures are low risk, they create an accessibility hurdle to a vast portion of the creditworthy population, particularly the young, the internationally mobile and in many parts of the world, the underbanked.

14. What is innovative about our product?

Social media is a highly personalized and internationally standard source of big data optimized for the millennial generation and, increasingly, the underbanked populations as well. Using it to for credit scoring is revolutionary for two reasons. First, we can analyze a lot of the same data that traditional scores do, such as education, employment and residential history, on a much larger portion of the population than traditional scores can. Second, we can find new predictive factors such as social network strength and behavioral, communication and lifestyle patterns, potentially improving the overall process of credit decision-making forever.

We plan to become the first international social media credit bureau servicing financial institutions.

Any Questions? Ask us!