Predictive lead generation is in and everyone wants to do it or are at least curious about it.
What is predictive lead generation?
Predictive lead generation is a strategic initiative to identify ideal customer profiles and discover net new prospects that fit that ideal customer profile. The difference between traditional lead generation initiatives that also are about net-new ideal prospects and a predictive initiative is that the latter is numbers-driven.
A predictive lead generation exercise should result in a rank ordered list of prospects segregated into segments.
Your predictive initiative is good as the data you have access to
Companies that opt for predictive lead generation are looking to answer the question “Which prospects have a better chance of converting into a customer?”.
One way to answer that question is by looking at past data and analyzing the profiles of your current clients during the time you acquired them and building an ideal customer profile. But where will you find this data? Your CRM tool will have some data. But it doesn’t capture all the data points. And whatever little data the CRM tool captures, it lacks the attributes and filters to single them out.
Let’s take an example.
You have a 15-year-old e-commerce logistics company which has acquired over 10,000 customers in the period of its existence. It’s likely that most of the sales reps who helped you win these accounts would’ve left the organization. You don’t know what the profile of that customer was at the time you made a deal with them. Were they shipping cross-border before they came to you? How many SKUs were they shipping?
A CRM tool won’t give you the complete picture.
Imagine having a platform which gives you the entire picture.
Over the last 15 years, you’ve interacted with 10,000 prospects. You converted 7,000 into customers. Out of the 3,000, you didn’t convert, 2,000 went till the last step.
Only if you’re able to analyze the past data clearly, you’ll be able to make a decent prediction about the future of the prospects in your funnel.
You are better at converting east-coast companies compared to west-coast ones. You have a better chance of converting companies that use this competitor. Most of your existing clients started cross-border shipping after you converted them.
These telltale signals will point you towards which prospect you have a better chance at winning. The CRM data alone won’t say anything. Enrich the data with custom-attributes and you’ll see the data shaping into a story. You’ll be able to see patterns emerging.
Any good predictive lead generation exercise starts with customer segmentation. Your prospects don’t fall into 2 categories as we’ve imagined. They can be classified into 8 categories. Based on past data, it looks like we have 3 high-value segments. We always thought it was one homogenous set!
The way you approach a lead-list/data provider would be different once you have a good handle on your customer segments. You can be very specific about your criteria and that save time, money and it dramatically improves ‘top of the funnel’ quality. Your MQLs will actually be of good quality.
Here’s an example of how specific you can get:
I want retailers and e-commerce sites that ship more than xxx SKUs every month and who already have XYZ packaging partner with warehouses in these areas in the US
I want companies with these many SKUs that already do cross-border shipping and use USPS. It should fall in this revenue slab.
Predictive lead generation in the absence of history
Predictive lead generation platforms will help you find companies you can go after based on companies you converted in past, companies you can pitch to and have a better chance at winning over as customers.
But what if you’re a new company? You won’t have past CRM data to learn from. What then?
In such cases, you depend on your team’s wisdom on what they think they can convert well. Instead of arriving at segments based on data, we use hypotheses to arrive at customer segments. Based on this you keep a close watch on your CRM and you’ll know what’s converting and what’s not converting. Wait for a year. You can slowly replace your hypotheses and guesstimates with actual data.
Your campaigns will have a higher success rate when you take into account the past data. When you don’t have any past data to learn from, you build models (Ideal customer profiles) and wait for a period of time before you can tweak it based on past data.