so_logo.png

Using Custom Test Types to Enhance the DOTS Lead Validation International Experience

DOTS Lead Validation International is Service Objects’ most popular composite service. It uses any combination of Name, Email, Phone Number, Address, IP Address and Business Name to help clients measure lead quality for the purposes of reducing fraud, increasing conversions and enhancing incoming leads, web orders or customer lists.

Lead Validation International uses our Name, Email, Phone, Address and IP Address validation services. blending in additional datasets and proprietary algorithms to return both an overall 0-100 score for the lead, as well as scores for the individual components. This gives users an idea of both how solid the component (such as Phone Number) is on its own, as well as how all of the components blend together to form a lead. In addition, many informational notes are added to give important details – both good and bad – about what we found out about the lead.

This blog dives a bit deeper into Test Types, what they are, and how custom Test Types can be used to handle specific (and sometimes unusual) client needs.

What is a Test Type?

A Test Type is one of the most important parts of Lead Validation International. It is essentially a set of instructions for how to process a lead. It tells the service which components (Name, Email, Address, Phone, IP and Business) to test, as well as which ones will be missing. Lead Validation International has hundreds of tests baked into the six main components. The score in positive or negative points can be controlled by a Test Type for every single test the service does. Weights can be applied to the components themselves (e.g. Name is 20% of the overall score). Thresholds can be set for Accept, Review, and Reject scores, both at the overall lead level and for each component, to give users a good idea of how strong the lead and each component is.

Test Types can also be grouped into 3 main categories: Residential, Business or Mixed Test Types:

  • Residential Test Types expect to see mostly residential data and penalize for data points tied to a business. They are designed for most B2C use cases. Matching is highly valued, so getting a few business data points does not fail a lead if things are matching up well.
  • Business Test Types are strictly expecting business data and penalize harshly for residential data points. Companies wanting to do B2B checks with clean business leads would use these test types (i.e. they negatively view residential data in these leads).
  • Mixed Test Types are really designed for B2B leads but value matching over the identification of business or residential data points. As long as the data points look solid and generally match up, high scores can be achieved with these test types. These are becoming the more commonly used Test Types, with more people working from home and providing at times residential contact points.

Limitations of Test Types

A large set of commonly used set Test Types can be found here:

https://docs.serviceobjects.com/display/devguide/LVI+-+Test+Types

These Test Types change slowly over time as new tests are added to the system, but they generally do not drastically change. They are consistently performing and attempt to adhere to expected industry standards for what is a good lead. However, not all clients might agree that a specific test should harshly penalize the score for a specific condition. But the expectation is that more clients than not will approve of the rule.

Test Types also strictly adhere to their purposes. For example, “business-noip” Test Type expects that the lead will be B2B with Name, Business Name, Email, Phone Number and Address. It would not give any points for the addition of an IP Address but would penalize a lead harshly for not having an address, for instance. To get around this, there is also a Test Type called “business-n-e-p” that expects Name, Business Name, Email and Phone but not Address and IP. These rules are important because a client can use “business-noip” to harshly penalize a lead that does not include a clean address. If they have mixed leads that sometimes do not have addresses, they can use the other test type for those cases to not harshly penalize the lead.

We have built out a large list of common test types to cover all of the various data input variations, however these remain pretty static and cannot be modified specifically for a client unless it makes sense to modify it for ALL clients. We will do that on occasion, and most clients share the same basic needs. Still, there are differences in needs that cannot be covered for all cases. A need for having an accurate lead in a CRM can be very different from a Marketing team allowing looser threshold holds to pass on a lead to a sales team. Some of this can be done via sliding adjustments on scoring, but in some cases fundamental differences in which tests to run can be problematic.

The solution to this is a powerful feature of Lead Validation International: custom Test Types.

How do custom Test Types help clients solve problems?

Custom Test Types are Test Types specifically tied to a client. Custom Test Types allow our users to control every aspect of scoring in Lead Validation International. They can control what data points will be expected, they can control the influence each data point has on the overall score, and they can control the Accept, Review and Reject thresholds of the overall score as well as the individual components.

While we do not openly publish our complete list of tests, clients can modify the bonuses and penalties of individual tests, and Service Objects will work with the client to make sure their needs are fulfilled. In an upcoming release, custom Test Types will be more malleable to handle missing data points. For example, if the custom Test Type expects Address but is also ok if address is not given, the Test Types will have alternative rules for handling those scenarios.

Here are a few common requests for custom Test Types, to give more specific ideas of their usage:

  • I want to fail any lead that contains a free email except Gmail (Hotmail, Yahoo etc).
  • I want to incorporate IP Address checks, but I do not care about location – I just want to look for danger flags (high risk countries, blacklisted IPs etc).
  • I want the lead to pass if either Email or Phone is good, and I do not care if the other is bad.
  • I want to move the overall Reject threshold from 55 to 45.
  • I want phone number to have a 50% influence on the overall score (commonly 35%).
  • I am more ok with business data points coming back in my B2C leads.

Custom Test Types are an extremely important part of the Lead Validation International service, and are becoming more and more common for our newer clients. We see client needs varying so much that normal Test Types no longer apply as often as they used to.

Custom Test Types are completely free to use for our clients.

Like the normal Test Types, we do not make drastic changes to custom Test Types, but for different reasons: custom Test Types should be owned by the client, so we avoid making changes that might affect their expected usage of that Test Type, but new tests may be incorporated if it makes sense for the client’s usage. However, we may reach out in cases of new tests and make sure the client wants to take advantage of these changes.

Hopefully this blog sheds some additional light on Test Types and how they are used in the Lead Validation International service. If you would like to learn more, feel free to visit our product page and developer guides or reach out to sales@serviceobjects.com to get in touch with one of our sales reps.

You may also like:

New resources, straight to your inbox

Get updates on the latest industry trends, tips, and news.