Score is everything that we know about a phone number, distilled into a single actionable value. Score gives a credibility level that you can attach to a given number that ranges from 0 (least credible) to 100 (most credible).
You should think about Score as giving you some useful guidance in how to trust the number that a new or potential customer is giving to you. We would recommend that a score below 50 should be treated with extreme caution and perhaps that you should perform some additional checks before you take it at face value, whilst scores over 75 indicate that you can be reasonably confident that the number is genuine and has a tangible level of both history and activity associated with it.
Only the mobile telephone number is needed. We don’t require name, address or any other personal information (although we also offer services that can check these too!)
No. When you invoke the Score API with a number we will provide you with the credibility score of that number, we are not necessarily guaranteeing that the person you are dealing with actually has that number. We strongly recommend that before you invoke the Score API you have already determined ownership of that number, either using a service such as TMT Authenticate or via your own means (such as an SMS One-time Password)
That’s the best part! Using our knowledge of global telephony numbering we are instantly able to spot incorrect or fake numbers, and return a score of 0 to leave you in no doubt that you need to ask the user for an alternative.
Depending upon the country, there are many different individual data points that are combined to give the Score. These include:
Yes. Although it should be noted that the score will be more developed and accurate in countries where there are larger amounts of telecommunications activity. Your TMT Analysis account manager. will be happy to explain this in more detail and provide guidance on which countries have the best data available to power Score.
The Score algorithm is a sophisticated set of instructions that adjusts the value of the individual data points based upon the real-world likelihood of them being seen. So for example, the fact that a number has not ported will have a different influence on the final Score depending upon the country it comes from, right down to having no influence at all in countries where porting is not currently available. Our developers are constantly working in the background based upon our experience to make the algorithm learn and Score customers take advantage of this continuous improvement without having to do any specific changes to the API.
As previously mentioned Score uses a variety of data points to come to a conclusion, some of them are on-net and very fast (less than 5ms) and some require off-net API calls to specialist suppliers and databases. All of this means that to get the most enriched Score result takes around 2-5 seconds to complete. However, if your business process can’t wait that long we are able to tune the algorithm to avoid the higher latency aspects of the data and respond faster with the on-net or near-net information. Whilst not the full feature-rich service, this is still adding considerable intelligence and value over doing nothing.
Given the ever-increasing importance of mobile in your customer engagement strategy, shouldn’t you be checking the authenticity of the number in the same way as you might about other information that you gather from potential customers? Although two numbers might look identical and correct, there is an inherent difference between one that has been active for 5 years with a steady and constant level of activity and one that had never been seen in the World before yesterday.
YES! Any learning algorithm such as Score is only as good as the information that it is given to learn from. TMT Analysis have poured everything we know about telephone numbers into Score but getting a customer perspective is always welcome. If you are prepared to securely share numbers that you know to be associated with fraud Score can use that information to identify patterns, improve overall detection and even allow you to predict high risk transactions based upon this in-depth analysis of history. The Score team are always eager to learn more and we would welcome further discussion on this.
Fornecemos os dados mais completos disponíveis sobre dispositivos, redes e numeração móvel
Contacta-nos > Conversa com um especialista >