First off, I want to provide perspective on my point of view around all things Predictive Marketing.
At Integrate, I head up data and technology partnerships, which means I’m out there all the time working directly with customers and prospects to understand ways Integrate can partner with other marketing technologies to provide a better solution. And in today’s B2B marketing landscape, to say that predictive comes up often would be a drastic understatement.
And while I can’t go five minutes without hearing “predictive,” it seems as if there’s no common definition of the term. This is a big concern among customers. And a common – or at least understood – definition is very much needed to assure internal stakeholders and external partners are on the same page.
It’s probably also important to point out the fact that Integrate sits in a unique space within the MarTech ecosystem; our software acts as a de facto execution arm for predictive platforms like 6Sense and Mintigo. This means I’ve had the privilege of working with predictive marketing pros extensively on both the technology and customer sides of the equation.
At this point, predictive isn’t new
Predictive Marketing is a concept that’s been around for more than a decade. With easy access to an abundance of data and emerging technologies, predictive data, modeling and analytics have become intriguing capabilities many marketers are trying to get their arms around. This marketing buzz has bubbled up various definitions of “predictive.”
Unfortunately, the broad, overuse of terms like “predictive analytics” and “intent data” has caused a significant amount of confusion for marketers and other pros who are seeking solutions that’ll help them develop more customers and revenue for less resources and investment.
The goals of predictive
Before we jump into the definitions, it might help to start with some common ground. For frontline marketers and data analysts, the goals of predictive is typically very simple: use a third -party technology and/or methodology to prioritize existing leads, targeting those with the highest likelihood of closing, and identify new leads that have a higher likelihood of turning into closed/won opportunities than those generated through traditional targeting methods.
That part’s pretty straightforward, but beyond this things get murky as the methods used to score those leads and identify those net-new targets vary greatly in terms of both cost and sophistication.
Key predictive terms
For the purpose of achieving the aforementioned common vocabulary, here’s our (Integrate’s) interpretation of some of the most frequently used language surrounding predictive marketing:
Predictive Analytics (PA) Lead Scoring: Technology that pairs your known CRM and marketing automation data with a variety of internal and third-party data sources, then plugs that data into a set of proprietary algorithms to attempt to predict which leads are in market for your product at a given time.
Predictive Analytics (PA) Net-New Targeting: Leveraging the same data, PA vendors will apply machine learnings to external data sources in order to identify prospects who are not in your database, but who are in market to buy your product.
Intent: Typically only indicated through predictive analytics, intent means that the lead/target has demonstrated behaviors that would signify they’re likely in-market and ready to purchase a solution similar to the one you’re offering.
Behavioral Data: Leads demonstrating a behavior (e.g., reading articles, downloading content, attending webinars, etc.) that would indicate some level of interest in the topic with which they’re interacting. The reason behind their behavior is unknown; therefore, behavioral data should be used as an additional point of intelligence to help prioritize leads, but shouldn’t be seen as “intent data.”
Look-Alike Modeling: A less sophisticated way of acquiring net-new leads, this approach involves analyzing your closed/won opportunities and identifying companies and personas that are similar. Methods for creating a look-alike model vary from company to company, but there’s typically nothing “predictive” (at least in terms of our definition) about the list of targets delivered to you.
This obviously isn’t an exhaustive list of terms. In fact, it doesn’t even begin to approach some of the more nuanced definitions surrounding predictive marketing. However, my hope in writing this is to create a core set of common definitions, we can discuss, debate and get agreement to advance predictive marketing’s impact on business.
With this core set of predictive terms agreed to, we can quickly move onto the more exciting discussions surrounding the tools, technologies and services that are going to help demand generation marketers and marketing ops pros achieve those elusive MQL, SQL, SAL and, yes, even closed/won goals.