Dun & Bradstreet Completes Acquisition of Lattice Engines. Learn More

Why Predictions Matter: Cloudy With A Chance Of Hot Leads

We often get asked the question – how is Predictive Lead Scoring different from traditional lead scoring? Why should we use a predictive model for lead scoring at all?


Let’s take an analogy. Let’s look at the weather. I looked outside about an hour ago. It was sunny. I look outside now, and it’s become cloudier. My guess is that it’s going to rain. I could go to the store around the corner and buy an umbrella to be safe. Or, I could just look at weather.com. 10 percent chance of rain! I think I will save my $10 for dinner instead.

So what did weather.com do that enabled it to say whether it was going to rain or not? In short, two things:

1. It used all the available data to figure out weather predictions much more accurately e.g., temperature in nearby towns, humidity, pressure, direction of winds, cloud patterns, etc. And it is just not data that matters; it is the trends in data that matter. Humidity by itself is important, but what is more important, is how humidity has changed over time.

2. It used a predictive model based on historical behavior to figure out the probability of rain. If in the past, the conditions that prevail now led to rain, it is much more likely that it will rain now.


This analogy applies to lead scoring as well. Lead scoring using rules is like looking outside and trying to determine the weather. Sure, it will work in several cases; if it is very cloudy and you see lightning, it will likely rain.

Predictive Lead Scoring is like going to your favorite weather app. Similar to weather prediction, we:

1. Use a lot more data than what Marketing Automation systems can provide. A contact might be active on your website, but we can provide account and contact level buying signals (e.g., growth indicators, key executive changes, social media activity, company financial metrics) that help determine whether the contact is a hot lead or not.

2. We apply a pre-built, pre-optimized predictive model that trains on your historical data, determines what factors are good indicators of hot leads specific to your situation, and use the model to determine which leads are hot.

For example, we found that growing companies are really good prospects for one of our customers. Through the Lattice Data Cloud, we monitor company hiring, revenue and other expansion trends that indicate growth and other buying signals that are not stored in Marketing Automation or CRM. Using this data, we were able to accurately predict the hot leads based on growth signals.

So next time you need to check the temperature of your lead – call Lattice.

Written by

Shobhit Chugh
October 21, 2013