Skip to main content
Radius Help Center

Best Practices for Outbound Marketing

Predictive acquisition helps you discover net new prospects that look like your best customers, and find other new prospects and markets. It can also help you uncover promising current prospects within your CRM. Read on to discover the science behind the search and get some guidelines for success using Radius’ predictive acquisition solution.

Understand the Principles and Processes

Radius helps you apply predictive acquisition to your outbound marketing and prospecting. Predictive acquisition is the process of using machine learning and predictive algorithms to identify the best prospects for your marketing or sales teams to pursue. Radius’ predictive algorithm works from two sets of data: your CRM and Radius' proprietary business graph, our comprehensive graph consisting of 50 billion dynamic signals on over 18 million U.S. businesses. 

The predictive acquisition process starts when you link your CRM to Radius. We match your records to the graph, uncovering information about each of your prospects and customers, including whether your previous marketing or sales efforts have been successful—or not. 

To help Radius discover promising prospects, you need to give us information about your marketing program, such as what criteria you use to identify successful and unsuccessful campaign results from the marketing channel you currently use, with similar future programs in mind. Knowing how you define success helps Radius make meaningful comparisons between records in your CRM and records in the graph, so we can optimize for the outcomes you want and deliver more good prospects. 

Let’s take an example. 

If your future goal is to convert Marketing-qualified leads (MQLs) to Sales-qualified leads (SQLs) and you want to use a direct mail campaign to do that, then the data currently in your CRM should reflect goals and channels defined the same way: prospects that have already converted (or not converted) from MQL to SQL from direct mail campaigns you’ve already run. 

The graph’s algorithm analyzes this information about your historical successes in a given channel, and translates those patterns it into a query that can search beyond your CRM to find similar new records that are likely to convert by the same means. And because the graph has richer data and deeper insights, we may also identify other net-new prospects you may not have considered, as well as existing prospects in your CRM that are worth retargeting. 

Know Your Predictive

It’s also important to understand the differences between predictive acquisition and predictive prioritization, or lead scoring. We’ve established here that predictive acquisition leverages historical success and customer characteristics for a given marketing channel to deliver promising prospects from your provider’s data set and your own. Predictive acquisition helps you smartly expand the top of your funnel. 

Another powerful solution, lead scoring, analyzes the same factors across customer and provider data sets, but it focuses on identifying the right prospects currently in your funnel: those you’ve already invested in, that much closer to Closed Won. 

Align Past and Future Marketing Goals and Channels

As we illustrated earlier, the historical data from the records you provide should align with your future goals, and your past and future marketing channels should also be the same. The greater the similarity between your historical success and your future goals, the better predictive results you’ll get. 

Understand Lift and Gain

Lift is predictive. It indicates how many times more likely you are to achieve successful responses by using our recommendations than you would be if you targeted prospects without them. So, for example, a lift chart might show that if you target 40% of your prospects using our recommendations, you’re two times (2x) more likely to get successful responses than you would be if you targeted 40% of your prospects  without applying our recommendations.

Gain shows the performance of a predictive model. It indicates the total percentage of successful responses you can expect to get when targeting prospects according to our recommendations as compared with random targeting. For example, a gain chart might show that if you target 30% of your prospects using our recommendations, the expected total percentage of successful responses is 85%—versus a 30% expected total successful response if you targeted your prospects randomly.  The more prospects you target, the closer the gain line gets to the maximum possible number of successful responses: 100%. 

Give Radius a Sizeable Data Set

For best success with predictive acquisition, it’s important to share a sizeable set of data, so Radius can identify patterns that are statistically significant and have more confidence in what they find and deliver. For our Segment Lift Scores feature, Radius requires a minimum number of records to compare (500 or more Won records, 500 or more Lost records, and 100 or more Open records), so if you’re new to your B2B market or light on data, you may need to do some hypothesis-driven acquisition first, so there’s something for the algorithm to learn from. 

  • Was this article helpful?