Exchange Performance Data and continous learning mechanism

Performance data is used to train our algorithms and hence improve the accuracy of our scorecard. By sharing data with FriendlyScore you ensure that our scorecard is continually optimised for your any borrower population. Data gathered is only used for your unique scorecard and not for those owned by other companies.

Data Flow diagram

Performance data, data flow standard

Database WhiteLabel Schema (how to update database to exchange Performance Data)

You can enter performance data directly into database. White label app will detect this fact and include this data to train the model.

If you look into the table “socialscore_user” you will see two important columns. One is “partner_user_id” which is your borrower identifier that you have entered to URL parameter when you have forwarded users to the White Label. Second important column is “wl_performance_data”. This column will be sent to FriendlyScore Learning Machine to improve the scorecard for your population. You can enter any string (and inform us how to interpret this data) but we suggest to put there a number which we can interpret automatically:

  • 0 - bad borrower
  • 1 - good borrower
  • 2 - borrower delayed in repayment

Or an extended status that will allow us to indicate a more advanced loan performance especially for installment loans

PaydayLoan & Recurrent Credit * 0 - No delay * 1 - up to 7 days delay * 2 - 7-14 days delay * 3 - 14-30 days delay * 4 - 30+ days delay

Installment Loan Statuses

  • 0 = No delay
  • 1 = Up to 14 days Delay
  • 2 = 14-30 days dealy
  • 3 = 1 month delay
  • 4 = 2 months delay
  • 5 = 3 months Delay
  • 6 = 4 months delay
  • 7 = 5 months delay
  • 8 = 6 months delay and more

Please notify us what will be your decision regarding the performance data status

Please update table “socialscore_user”, column “wl_performance_data” for specific “partner_user_id” to have best accuracy on scorecard for your population of borrowers.

Other posibilities to exchange data

We have a team of dedicated software engineers who can help channel performance data to FriendlyScore. We are happy to appoint a member of our team to support

Data Connector which can be installed on your website to easily exchange information between systems.

To request FriendlyScore Performance Data Connector please send the information below to

  • Company name / Scorecard ID
  • Operating system on server (Windows/Linux/Other)
  • Database schema with tables where performance data is stored on your server
  • If you do not store performance data about your customer, please describe how you interpret good/bad customer performance from the database schema that you send to us

We are open to write dedicated software for selected FriendlyScore partners. If you want to become a partner please write to As a partner, our developers will work directly with your team to create a tailored solution for information exchange.