Transactional Monitoring in today’s world is necessary to implement variable plans and tiers of services to our customer base. With intelligent systems being put into place at every junction of the transactional process from payment till the evaluation of end numbers of purchases, we sought to find an intelligent source of digital data that can drive insight generation like no other in the market.
With our continuous pursuit to help businesses reduce the blind spots in business profiling metrics, we came up with yet another feature that will prove to be the richest source of financial data ever in a system for transaction monitoring, i.e SMS TxN Extractor. SMS was believed to be a dead market and an ancient technology that no one pays attention to anymore, but recent years during the pandemic we began to understand the reliability. Apart from the concept of reliability of SMS messaging, texts are quick, cost-effective and results can very quickly be analyzed through OCR or even more quick basic tools. SMS Txn Extractor filters the transactional SMS’s from collection of SMS’s that we receive every day, to evaluate the user attributes that will be helpful to predict many different vectors such as – Purchase pattern, Financial profile, Income proof and Credit/insurance risk.
First, here’s a short roundup of the benefits of using SMS TxN Extractor for your business and what it can do, after which we’ll see how the monitoring of transactions are actually useful in drawing informed conclusions in day-to-day business decisions!
How would this help in transaction monitoring?
1) Helps one understand the purchase pattern: Transactional SMS captures beyond overall debit and credit. It can tell you about the usage and frequency of purchase of various products and services by extracting SMS’s from E-commerce, App-based cab, Mobile wallet, OTT etc.
2) Helps you understand the financial profile: Transactional SMS is a rich source of information. It can tell you regarding earning and spending patterns.
- It’s a tedious, manual, and erroneous task to segment actionable financial profiles from bank statements.
- Information on transactional SMS: Salary, cash withdrawal, utility expenses, e-commerce expenses, bank balance, number of bank accounts, delayed payments, etc.
3) Gives you information regarding credit & insurance risk vectors: Transactional SMS covers many more indicators beyond credit card and loan repayment delays.
- Cheque bounce is not covered by most credit rating agencies, but SMS extractor captures this.
- Delay or default in e-commerce, mobile bill, and other utility payments are not captured by credit rating agencies but SMS extractors can extract that information as well.
4) Inputs about the Income proof: NLP of transactional SMS provides real-time passive way of doing income proof for 100% of the population with bank accounts/mobile wallets.
- Only 23% workforce is salaried and have income proof document.
- Present income proof is mostly offline and a tedious process with high fraud risk.
- Real-time bank statement checks with tools like Perfiose has 70% plus drop rates due to lack of trust in giving critical information also Perfiose can only give a view of the account the person has logged in.
How do we implement transaction monitoring ?
- The transactional SMS’s from a whole lot of messages are filtered out.
- Virtual passbook is created by applying ML algorithms to the messages. SMS transaction extractor solution applies its intelligence to categorize the messages into various expense types across products & services, purchase patterns, and credit & insurance risks.
- Duplications are removed and filtered out.
- If there are any incomplete values in the data, then they are filled.
Percentage of Subscribers having different transactional messages –
Provision of displaying multiple bank Accounts-
Try IDcentral’s 6-month free ID verification solution.
Sushanta Mishra is Associate Director at IDcentral. He has around 14 Years of experience in Banking, Insurance & B2B industry. He has worked on numerous transformational projects across Machine Learning & Deep Learning.