When it comes to lead generation, quality is more important than quantity in the long run.
But when you have a long list of prospects, how do you predict which are potential buyers and which are a waste of time?
We recently spoke with Tony Yang, VP of Demand Generation at Mintigo, who explained how predictive marketing can help with the sometimes sticky task of lead qualification.
Predictive marketing helps you determine whether or not a lead or account is a good target to focus on. More importantly, it tells you why.
It uses data modeling, which is taking a jumble of unstructured data collected from across the internet and plugging it into a model, so it becomes usable statistics that predict whether a lead will become an actual buyer.
Tony breaks down the how, what, and when of predictive marketing for us. Here is what he shared.
How Does it Work?
The idea of data modeling is great, but it doesn’t tell you much about what is actually happening within a predictive model.
Basically, the model builds an ideal customer profile and then the program does the work of figuring out if prospects match up enough to warrant your attention.
To do that, you want to start off with a “positives list”. These are your current clients: the people and companies who have already purchased from you.
The model is looking for insights into common characteristics of your customers by comparing a variety of data points. These common characteristics are your customer DNA or ideal customer profile.
Once you have that in place, you’ll build a “negatives list”. These are the prospects or the unknowns.
When both the positive and negatives lists are run through the model, it gives the prospects from the negatives list a score based on how closely they match the characteristics from your customer DNA.
That score is a comparison between the prospects. With this information, you can easily see which prospects are more likely to buy, which means you can focus more of your attention on winning them over.
What Types of Data Should You Look For?
There are a variety of data types that can be used in a predictive model to build out your customer DNA.
For instance, what type of technology do your clients use? Are they on Salesforce? Microsoft Dynamics? An ERP system?
Do they use some sort of database tech that will make them more likely to buy your product?
What are their hiring trends and patterns? What sort of organizational roles do they have?
If it’s a public company, is there any financial data that can be used by the model?
You can also look at intent data. This is activity that a company or individual exhibits across the web that points to potential buying signals.
For example, a marketer from Company X is looking for new SEO tech, so they’ll probably do research online to find what fits them best. On the other hand, a marketer from Company Y who isn’t looking for SEO tech could be reading the same types of articles just to keep up with the times.
If the data is run through a machine learning to look for true signals of intent, these two hypothetical people can be differentiated in the system.
It’s unlikely that you will have access to all of these types of data points within your CRM system. That’s where data collection from predictive vendors like Mintigo comes in handy.
When do You Have Enough Customers to Benefit From Predictive Marketing?
There’s not really a hard number for this.
Some companies start around 400 current customers, while others could find it useful with as little as 200 customers.
It depends on the needs of your business.
That said, if you have five customers, predictive marketing is going to be less helpful than just calling your customers and interviewing them to gather qualitative data.
Having that small of a data set just wouldn’t give you significant information for a predictive model.
Still, if you’re in the less than 400 range and think predictive marketing could help you cultivate better leads, there is a workaround that you could use.
Instead of focusing solely on current customers for your positives lists, you could supplement that list with late-stage opportunities. These prospects haven’t purchased yet, but if they’re far enough into the sales funnel that it seems likely they will become customers, the data could be just as helpful.
With predictive marketing, you can accurately predict the quality of a lead instead of just guessing if they are a potential buyer.
Creating positives and negatives lists help slim down the vast pool of leads that can be overwhelming. It helps you prioritize accounts and focus your energies on leads that will turn into actual buyers.
Predictive marketing takes “knowing your customer” to a whole new level.
This article is based on an interview with Tony Yang, VP of Demand Generation for Mintigo.
Listen to the episode that this post was based on here:
James Carbary is the founder of Sweet Fish Media, a podcast agency for B2B brands. He’s a contributor for the Huffington Post & Business Insider, and he also co-hosts a top-ranked podcast according to Forbes: B2B Growth. When James isn’t interviewing the smartest minds in B2B marketing, he’s drinking Cherry Coke Zero, eating Swedish Fish, and hanging out with the most incredible woman on the planet (who he somehow talked into marrying him).