The ongoing battle against robocalls is multilayered. User-level robocall mitigation identifies and blocks spam and scams automatically through an app. Savvy services might play an “out of service” message so that unwanted robocallers remove the number from their call lists to deter future attacks. This approach is effective for entrepreneurs and small businesses as well, since such operations often consist of one or a handful of users. For larger enterprises, we talk about brand reputation management. When bad press leads to lost customers, detecting and eliminating imposter traffic, protecting brand image, company reputation, and customer relationships is critical. To enforce brand reputation management, enterprises employ web scanning tools to detect misuse of logos and domain names, as well as malfeasance over social media, email (phishing), and voice services (vishing). Another wrung up the robocall mitigation ladder lies in fortifying the voice networks themselves over which fraudulent activities originate and traverse. This is the layer in which analytics, attestation, and artificial intelligence (AI) take center stage.
In this article, we cover:
Nearly 50 percent of calls are unwanted. In parsing nearly 4 billion robocalls received nationwide in May of 2022, YouMail found that 56 percent were categorized as scams or telemarketing spam, 28 percent were alerts, and the remaining 16 percent were payment reminders. That averages out to 1,500 robocalls per second. While it may sound overwhelming, the lemonade that can be pressed from these lemons is data — raw data that leads to useful information.
There are two particular ways in which attestation software leads to powerful conversational AI: event-based analytics and content-based analytics.
Event-based analytics relies on call-event data to predict call outcomes. An algorithm leverages network signaling such as SIP messages, answer seizure rates, call hold times, and more to assess the likelihood of spam or fraud. However, such predictions as intent can only be proven post-fact.
Content-based analytics, on the other hand, parses actual telephone call content to understand what has occurred. Did the caller self-identify? Did they provide an opt-out number? Does the known legitimate entity engage in voice communications at all?
Attestation software that leverages a combination of event and content-based analytics can provide the right mix of deterministic and probabilistic data necessary for an extremely high degree of analytical veracity. For example, if your tools understand the probability that millions of called parties all have the exact same charge on their Amazon account and that the online retailer does not place certain types of calls in the first place, your communication service has found the alignment of certainty it should seek.
The most powerful solutions are capable of algorithmically scoring telephone numbers. Content-based analytics should be able to provide a probability of spam, unlawfulness, or fraud. YouMail Protective Services customers and partners — including communication service providers (CSPs), enterprises, and third-party vendors — rely on YouMail PS Score for that exact kind of informed decision making.
For greater traffic analysis, it is best to couple CSP data with content-based analytics. YouMail PS is a fully integrated network solution that analyzes every communication instance handled by a CSP with comparative analysis to data (including audio content) captured by the YouMail Sensor Network, which consists of 350 million consumer and small business phone numbers mitigating over 1 billion robocalls out of 10 billion calls received.
YouMail PS provides visibility into usage behaviors which serve as an excellent complement to Know Your Customer (KYC) processes and procedures. Service providers use our scoring for real-time decision-making, including labeling, redirecting, and/or blocking calls. The YouMail PS Watch data dashboard supplements service provider KYC with a convenient and easy-to-use web-based portal that provides a visualization of unwanted robocall campaigns and content details. Our solutions offer powerful conversational AI algorithms to identify telephone numbers that are originating unlawful calls, and our threat database maintains a continuously updated list of telephone number and their associated campaigns.
Every CSP needs event-based analytics and content-based analytics to successfully predict outcomes based on an assessment of ALL data presented. If you’re ready to reduce your risk from FCC fines and actions, schedule a demo with YouMail PS today.