Revolutionizing Credit Scoring in Tier 3 Cities and Villages: The Social Media Solution

In a rapidly changing financial landscape, traditional credit scoring methods can be limiting when it comes to serving underserved communities in tier 3 cities and villages. Enter the innovative approach of leveraging social media data to evaluate the creditworthiness of individuals with little to no financial digital footprint. This blog explores the journey of one of the biggest banks in Indonesia, where they tapped into the world of Instagram and TikTok to bridge the credit gap.

Challenges: A Blank Digital Slate

In the vast landscapes of tier 3 cities and villages, individuals often lack a substantial financial digital footprint. This scarcity of traditional data makes it challenging for banks to assess the creditworthiness of potential customers. Without a credit history or any substantial records, how can financial institutions make informed lending decisions?

Opportunities: The Rise of Social Media

While these individuals might have minimal financial data, they are far from being digitally inactive. Many of them are highly active on social media platforms like Instagram and TikTok. This digital activity opened up a new world of possibilities for assessing creditworthiness.

The Solution: Leveraging Social Media Data

The Bank decided to explore this untapped resource by requesting access to the Instagram and TikTok accounts of potential customers through API-based access. This initiative marked a significant shift in the way credit scoring is traditionally done. Instead of relying solely on financial history, the bank began to analyze users’ interactions on these platforms.

Mapping Users: Low, Medium, High Categories

The bank developed a robust algorithm that evaluated users based on several factors:

  1. Interaction with Users with Valid Credit Scores: Users who engaged with individuals known to have good credit histories were assigned higher scores.
  2. Engagement on Social Media Platforms: The frequency and level of engagement on Instagram and TikTok were analyzed to gauge the users’ level of social activity and connectedness.
  3. Content Consumption: The type of content consumed also played a significant role. Users engaging with content related to financial literacy and responsible financial behaviour were considered positively.

Accuracy Achieved: 70%

Through this innovative approach, The Bank was able to accurately categorize potential customers into low, medium, and high-risk groups with a remarkable 70% accuracy rate. This was a groundbreaking achievement, given the lack of traditional credit data.

Dynamic Credit Scoring

What sets this approach apart is its adaptability. The Bank didn’t stop at the initial categorization. Instead, they continued to monitor users’ social media activities, allowing them to adjust ratings and categories accordingly. This dynamic approach ensured that customers’ evolving financial behaviours were reflected in their credit scores.

Pseudo Code for Social Media-Based Credit Scoring

Let’s break down the pseudo-code for the social media-based credit scoring system:# Import necessary libraries
import requests
import social_media_analyzer

# Define user’s social media accounts
instagram_username = “user123”
tiktok_username = “user456”

# Get API-based access to social media data
instagram_data = requests.get(f”{instagram_username}”)
tiktok_data = requests.get(f”{tiktok_username}”)

# Analyze social media data
social_media_score = social_media_analyzer.analyze(instagram_data, tiktok_data)

# Evaluate user’s creditworthiness
if social_media_score > 0.7:
credit_category = “High”
elif social_media_score > 0.4:
credit_category = “Medium”
credit_category = “Low”

# Update user’s credit rating and category in the database
database.update_credit_rating(user_id, credit_category)

# Continuously monitor and adjust credit scores over time
while True:
new_social_media_data = requests.get_updated_data(instagram_data, tiktok_data)
updated_social_media_score = social_media_analyzer.analyze(new_social_media_data)
if updated_social_media_score != social_media_score:
database.update_credit_rating(user_id, updated_social_media_score)
social_media_score = updated_social_media_score

This pseudocode outlines a simplified version of the process. In practice, the algorithm would be more complex and involve extensive data analysis.

Conclusion: A Bright Future for Inclusive Banking

This Bank’s pioneering approach to credit scoring has demonstrated that the world of social media can be a powerful ally in extending financial services to underserved communities. As technology continues to advance, financial institutions worldwide may look to leverage alternative data sources, like social media, to create more inclusive and accurate credit scoring models. This innovative approach has the potential to reshape the landscape of banking, making financial services accessible to a broader range of individuals, regardless of their traditional credit history.