How to Choose the Best Reverse Phone API

There are a lot of solutions in the marketplace for reverse phone data. We have been in the phone data space long enough to know how to identify the best solution, and we wanted to share some of the insights we have gathered. When vetting reverse phone API solutions, it’s important to understand the data behind the products, have a good framework for assessment, and look at some technical considerations to make–all of which are covered below.

Types of data: A good place to start a vendor conversation is to ask, “How do you source your data?” Most solutions leverage a single type of data, but there are a variety of data types included in reverse phone API products. Here are the categories of data you should expect to hear: 

• • Phone validation: The most basic form of a validation check ensures that a phone number includes the correct number of digits and syntax (e.g. area code + 7 digits). There is also something called the North American numbering plan, which allocated phone numbers to certain regions and carriers. This data determines if a number is valid based on where it has been assigned to a carrier. In other words, a phone number that was never assigned to a carrier can’t be valid. One thing to note: this is not the same as whether a phone number is actually in use. 
  • Telco metadata: 
    • Metadata: There is a network of data available to provide insights about the ported carrier and line type data included in the response.
    • CNAM: This is a telco-based service that provides the 15 character string associated with the billing name of an account (typically the primary account owner)–just like you would to see on landline caller ID screens. Carriers have largely moved away from providing direct access to more robust account holder information.   
  • Licensed identity data: There are many sources of licensed data out there. They can be based on public records, utilities, scraped data, credit-based, etc. It’s always good to know the providence of the data to ensure it’s reliable. These can be accessed via files or APIs.

How to evaluate: In order to understand the efficacy of any data solution, it’s important to assess both coverage and accuracy. There is a delicate balance between the two. Totally inaccurate data with full coverage is pure noise, and highly accurate data that is rarely available is often less practical to implement. Therefore, assessing both in your evaluation is key. 

• • Coverage: what percentage of the time we are able to associate information with a given phone number. Coverage can vary widely depending on the attribute you’re evaluating. For example, here is a range for typical coverage rates:
    • Phone validation: 100%
    • NPA-NXX based location: 100% 
    • Name coverage: 30-90%
    • Full address coverage: 0-89%
  • Accuracy: 
    • The first accuracy check everyone runs is their own phone number. This is a logical step, but it’s important to not stop there. Create a sample set of data based on friends, family, colleagues, etc. that you can use for comparison. 

Technical considerations: Today, enterprise-class delivery is table stakes for API-based solutions. Here are a few specific areas of focus:

• • Latency / throughput: It’s important to understand the speed of the service you are evaluating, which is typically measured in milliseconds. For any kind of real-time lookup, it’s important to target the fastest response time possible; I’d estimate something in the range of 30-500MS. If you run batches or frequently have large queues built up, it will be important to understand what kind of rate limiting or queries per seconds (QPS) constraints there are on a key. 
  • API response: These days, most APIs are RESTful, http-based and return JSON responses. Understanding the response set, documentation and support infrastructure, its consistency in terms of values returned and how the company thinks about any breaking changes is important for any Enterprise grade APIs. 
  • Service Level Agreements (SLA): Knowing the reliability of the service is key. With cloud infrastructure providers like AWS and Azure, keeping high levels of availability is the expectation. You may also want to know what kinds of alerting and monitoring the API provider offers. 
  • Privacy and Security: As this is PII data, both from the business and the queries being made by the customer, it is critical that the appropriate privacy and security standards are applied to safeguard this data.

Whether you are vetting your first data solution or you’re an experienced data provider, you should keep these factors in mind when choosing the best reverse phone API. If you’d like to take a more in-depth look at data sourcing at Trestle, take a look at this recent blog we published.