Light bulb We've updated our Privacy Notice. Please click here to see a summary of the changes.

Skip to content

Fight Account Fraud With Phone Number Reputation Scoring

Telesign Team June 15, 2021
person holding a smartphone with low risk meter juxtaposed

So, you’ve enabled two-factor authentication (or 2FA for short) and think that’s enough to protect to protect against fraud? Think again. The explosive global growth in using digital services, along with increasingly sophisticated identity theft methods, have exposed the limitations of relying solely on 2FA.  

This is where phone number reputation scoring comes in. On a high level, it’s the process of assigning a “score” to a phone number, which can help weed out numbers used for fraud from real ones. Phone number reputation scoring is critical for validating customer identity, preventing fraud, maintaining community integrity, and creating a better user experience.

Why Phone Number Reputation Scoring?

Fake online and mobile app accounts are being created at an alarming rate, resulting in spam, phishing attacks, promo abuse and other costly fraud. When new users aren’t properly verified, companies risk allowing bad actors onto their platforms. This leads to user abandonment, negative brand impact, and decreased revenue and growth. Identifying and blocking harmful users at account registration, while streamlining the process for real users, is vital to building a robust platform for customers.  

On its own, 2FA isn’t a failsafe to prevent account fraud. Using a phone number for new account registration only verifies existence and possession (which are two common authentication factors), but it doesn’t check to see if the new account itself is risky. With phone number reputation scoring, several types of verification signals are analyzed, including:

  • Traditional identification data sourced from wireless carriers like contact match, location, activation/deactivation, and porting history
  • Application-to-Person (A2P) data that evaluates activity with digital assets like commercial and financial web properties
  • Person-to-Person (P2P) communication data like messaging and voice data, successful calls completed, call durations, and transaction tenures  

Taken together, this information helps build a more complete risk profile for each phone number used to create a new account. The most relevant risky behavior associated with a phone number is identified and helps companies accept or reject a user based on their reputation score before sending a 2FA verification, such as a one-time passcode via text. Reputation scoring has been shown to decrease fraudulent activity by as much as 40%.

What to Look for in a Phone Number Reputation Solution

When evaluating what phone number intelligence to add to your fraud platform, there are a number of factors to consider:

  • Static phone number data: At a minimum, a successful solution should be able to determine whether a number is a VoIP number (which is inherently riskier than a mobile number) or an invalid number.  
  • Global capability: Fraud activity can come from anywhere and varies based on geographical location. It’s essential to analyze users’ phone numbers and associated behavior on other online services, including websites from around the world.  
  • Real-time risk evaluation: The Internet never sleeps. Getting real-time risk evaluation data cross-analyzed from several domains indicating irregular online registration activity helps flag risky phone numbers that are being used across multiple platforms.  
  • P2P communication data: Identifying usage and fraud trends from messaging and voice traffic is important to provide a more accurate and complete risk profile.
  • Cross-customer fraud & telco data: Leveraging reported suspicious network activity data will identify scammers quickly and stop them from impacting your ecosystem with fake accounts and fraud.  

Every business is different. The ability to customize phone number intelligence data and easily integrate a machine learning model into your larger fraud platform will help you optimize your fraud security practices.  

How It Works With Telesign

During new account registration, a user enters their phone number. That phone number is passed to Telesign to for evaluation. Telesign’s Score API uses machine learning to evaluate phone data and user activity attributes identifying any potential red flags. Score returns a numerical risk assessment, fraud score reason codes, and recommends whether to accept or block a number or if additional information is needed to make a determination.

Enterprises can customize reputation score threshold levels based on business risk models to deny or accept a new user or integrate Score API into their fraud management systems to provide a holistic risk evaluation.

Not All Phone Number Intelligence is Created Equal

As an industry leader in 2FA and digital identity security, Telesign has mastered advanced phone number intelligence. Our access to an extensive network of insights on billions of phone numbers and communication behavior metadata is unparalleled, all thanks to our working relationship with the largest international web properties, our cross-customer fraud intelligence, and our association with BICS and its global telco fraud database.  

Telesign can analyze both A2P and P2P activities, including phone number usage velocity, traffic patterns, history of fraud, phone type, telecom carrier, subscriber contact details, and status of phone, to determine appropriate risk levels in real-time. Our machine learning (ML) process uncovers hidden insights in the data to come up with a range of predictors for both good and bad phone numbers.

What’s missing from other phone number reputation solutions is the communication behavior signals that Telesign provides, which are sourced from our extensive global phone number intelligence network. Other vendors only provide traditional phone intelligence data from wireless carriers they have existing relationships with. This data can provide a risk or trust score, but you might be missing other relevant fraud activity that could be used to determine the full business risk profile of a new customer sign-up.

Understanding call behavior, phone number range data, number activity, transaction duration and other P2P communication data signals helps you identify ongoing user risk patterns to catch fraud with better accuracy. Fraud is everywhere, but having the ability to globally look at all phone number intelligence scenarios is the key to knowing your customer.

Contact us today to find out more about phone number reputation scoring and filling the gap in 2FA.