Imagery Majestic - Fotolia


If you need to close more deals, consider predictive lead scoring

Today, companies struggle to find the most promising sales leads. Predictive lead scoring can make the process more efficient.

For many companies, today's marketing and sales funnel is a confusing mess. The top of the funnel is swelling with increasingly more leads that are responding to an avalanche of content marketing and freemium, or try to buy, offers, while the bottom of the funnel can seem like a trickle, as salespeople struggle to identify the most promising leads, fueling lower sales rates, lower average selling prices and longer cycles to close deals. 

The marketing and sales funnel encompasses both marketing and sales processes, from a company's initial contact with a prospective customer to the final sale. Marketers are responsible for the top of the funnel, working to gather contact information and basic qualification data for new sales prospects. As it narrows toward the bottom, the funnel represents sales prospects who are further along the buyer's journey, or prospect lifecycle. The funnel can become a flashpoint between sales and marketers when marketers provide sales with poor-quality leads and expect sales to close deals nonetheless; and when sales fails to follow up on good prospects in a timely manner, leaving the marketer's work fallow and wasted.

For many companies, marketing automation platforms' (MAPs) rules-based lead scoring does not go far enough to fix the problems with the funnel. Last year, Sirius Decisions found that 68% of companies used MAPs to do lead scoring -- but only 40% of salespeople believed that it was effective. That's why companies are turning to predictive lead scoring, which harnesses the power of big data to identify the leads that are most likely to convert, close and generate the maximum revenue for the company.

The hard numbers behind predictive scoring

Predictive lead scoring can provide a "credible, third-party source of truth between sales and marketing," said Melissa Davies, head of global marketing operations at SLI Systems, which uses Lattice Engines, a predictive lead scoring technology. When salespeople focus on the top leads, sales can grow. Software vendor Infer has helped ZipRecruiter salespeople to spend 90% of their time getting high-quality leads compared with just 72% in the past.

By focusing on the right leads, DocuSign, which also uses Lattice Engines, experienced a 38% increase in the predictability of conversions, said Ryan Schwartz, director of marketing systems and operations. Lattice Engines' CMO, Brian Kardon, said that predictive lead scoring can shorten sales cycles from 180 days to as few as 100 days, and can increase average deal size by 20%. Finally, continuous success metrics and automated models make it easier for companies to refine and update lead scoring to increase accuracy over time. So, how does predictive lead scoring work?

Predictive starts with the model

The first step is to develop a predictive model based on a company's particular needs. The good news is that you don't need to hire an in-house data science team to create the model. Within a few weeks, cloud-based predictive lead scoring vendors can provide a learning database to create an automated lead scoring model. The database incorporates the customer's lead and won/closed data, and other data from ERP, MAP, CRM and customer service systems, as well as external information.

The vendor's software then creates an automated model that identifies the best positive and negative success predictors from thousands of potential internal and external signals and attributes. Some customers will hold back a portion of their data for back testing in order to assess the model's accuracy. Typically, you need 100 successful outcomes (e.g., 100 wins) to create a reliable model, but the vendor can provide guidance based on its experience and your needs. So, predictive lead scoring probably won't help you if you're a very new company with few successful outcomes or wins. 

A score itself is neither actionable nor sticky: It's what you do with it that counts.
Vik SinghCEO, Infer

The vendor's analytical engine then uses machine learning to identify the top success predictors, and creates a predictive scoring algorithm based on the relative weight of each attribute. The result is a straightforward lead score that might range from one to 10, indicating a low-to-high likelihood to buy, lifetime value, and/or stage of the buyer's journey. Some vendors also provide a rationale for each score.

As new leads and new data come in, predictive models send lead scores to your MAP and CRM systems using prebuilt application programming interfaces (APIs). "The ideal lead scoring system should show the 'why' versus a black-box approach,” said Alison Murdock, vice president of marketing at 6sense. "The 'why' tells sales reps what buyer activity led to the score so the rep can use the intelligence to drive a better-informed sales process."

Shedding light on the black box

Speaking of black boxes, the software can lack transparency about some of its assumptions in lead scoring and other analysis. These technologies use algorithms that can be difficult to work with in terms of these assumptions. Many companies want greater transparency -- or control of -- how these algorithms are set and what the criteria and priorities are.

The black-box nature of these models creates another challenge: They may not be intuitive to, or believed by, stakeholders. The whole idea of predictive lead scoring is to identify the factors -- some expected, some not -- that drive positive results. But just like any insight, predictive lead scoring requires humans to correctly interpret the data (i.e., to make sure that the model is accurate). And that means bringing in not only marketers but also sales leaders and potentially even independent data scientists who can assess the model's validity.

Continuous improvement

Expect to develop the model iteratively to improve accuracy -- both before and after go-live.  By using an iterative approach to model development, DocuSign's Schwartz emphasized, the focus is on the accuracy of the score. Although the lead scores should be reliable, they may also be counterintuitive to sales reps. So, you may want to keep the lower-scoring leads in marketing for nurturing, passing only the high-scoring leads to sales so they can focus on closing deals with the right leads. "Developing accurate scoring is important, but what's necessary for achieving success is the application of the score," said Vik Singh, CEO of Infer. "A score itself is neither actionable nor sticky: It's what you do with it that counts."

Predictive lead scoring systems also include reporting that enables marketing and sales leaders to track the effectiveness of the models, so they can be modified as needed. As you add more data, the model can become smarter and may be updated on different time frames, depending on your needs.

Knowledge is power

As I mentioned in my last article, knowledge helps your salespeople to more effectively prioritize, sell and close business. Predictive lead scoring can deliver the intelligence you need to do just that.

Next Steps

Lead generation software offers sales new tools of the trade

Filling intelligence gaps with marketing automation

Knowledge automation and clickable intel

Dig Deeper on Marketing automation