Lead Scoring: Everybody’s Doing It… But You May Need an Upgrade

Is everybody doing lead scoring? In some way, yes. Some are using predictive analytics. Others are using a business rule- or points-based approach, often leveraging the built-in functionality in their marketing automation tool. Others still are simply drawing on intuition and experience to score and prioritize who they call on or slot into the next outbound campaign.

Ultimately, lead scoring is a tool for sales and marketing teams to prioritize investment in marketing campaigns and sales plays and communications. Regardless of approach, lead scoring allows organizations to bucket their leads into groups (hot | warm | cold… high | medium | low… you name it). It’s all about having a manageable way to prioritize. Why do it? Simply put, lead scoring drives financial outcomes (higher revenues) but also other key interim key performance indicators such as pipeline velocity, lead qualification accuracy, sales costs, and close rates.

Best Practices For Lead Scoring

We believe that there is a progressive approach to maximizing how that prioritization happens and leads are scored. The right level of maturity is a function of return on investment and your organization’s ability to implement, adopt, and surround your lead scoring with other key pieces that will make it successful. That said, a great New Year’s resolution for all of us is to identify where we are on that roadmap and upgrade our approach to lead scoring, no matter what group you’re in.

Group 1 – Moving from intuition to data:
The case for making data-driven lead scoring decisions has been stated for more than a decade. Marketing Automation providers have helped pave the way with loads of statistics and case examples that make this a no-brainer to test (192% improvement in lead qualification according to Eloqua and 22% higher likelihood of sales teams hitting quota from CSO Insights – via Marketo). The analysts at SiriusDecisions and Forrester overwhelmingly agree.

So, how do you get from an intuition-based approach to a data-driven solution? If you have a marketing automation tool, the job is a bit easier as there are GUI-driven tools to help guide the creation of your own lead score. But it is also easy to do in your CRM tool or a spreadsheet too. Start with defining the fields that you want to use to create a score. We typically start with general account level attributes like firmographics and layer those with behavioral attributes available at a contact or buyer level. The former provides a good baseline for how you fit, the latter is indicative of interest and readiness to engage and possibly buy. Once the fields are define, assign the relative “points” (or weights) that you want to give each field. Add them up and, voila, you have a lead score. Once every lead has a score, you simply rank order them and group them into categories that help you easily prioritize. Even if your selection of which fields and their relative points isn’t perfect, you’re using data instead of gut instinct to prioritize, which is the right first step.

Group 2 – Moving from points to predictive lead scoring:
The research here is fairly strong in support of the opinion that predictive solutions are, on average, more effective than the alternatives. According to SiriusDecisions, 90% of those using predictive lead scoring agree that it is better than traditional approaches. Granted, this is a biased sample (as they’ve had taken the leap for a reason) but one where most of the users have used both and are seeing real returns in the shift to predictive models.

If you’re using a business rule or points based solution, you are much of the way there, believe it or not. Predictive analytics help us define two things well in order to prioritize leads: what fields are meaningful and what weights or points are applied to each. This are the exact two things that drive a points based process. The primary value that predictive statistics adds is a more data-driven approach to making those two decisions.

Once those fields and weights are defined, the creation of the score is effectively the same. Building the initial predictive model does require some statistical skill and a tool. You can either build or buy the skill and easily feed your scores into the tools you use every day (probably your CRM and Marketing automation tools). Alternatively, you can layer in a separate piece of software that focuses on lead scoring.

For more thoughts about the mechanics of predictive analytics and use in lead scoring for sales check out our whitepaper titled: Help Sales Close Deals Faster with Predictive Analytics.

Group 3 –Better value from existing predictive lead scoring:
For those already using predictive analytics consider how to improve your outcomes. This comes in a few forms:

  • Better data inputs feeding the predictions: This includes additional sources, third party data for enrichment, and better quality and cleanliness of the data that you do have
  • Improved adoption: More to come on this topic later but it is critical that the predictive scores you have are used and trusted. Reps and marketers can worry that predictive-based lead scoring is black box and less accurate so it’s critical to provide visibility as to the key fields that drive the prediction and educate users (ideally with positive testing outcomes as a known result).
  • Better support tools: It is important that the scores don’t stand alone. The right content and messaging are critical pieces support a personalized experience and make your lead prioritization effective.
  • Enhance methodology: Lastly, dig into your methodology –for those who are already effective at the points above—it is important to consider if you want to get to the next level. Drawing on more sophisticated statistical approaches to improve accuracy and, specifically, moving toward a machine-learning based approach will move the needle. With machine learning, the predictions are learning and optimizing constantly. As leads engage and buy or their characteristics evolve, the system logs those new characteristics and outcomes. The model takes these new data and effectively tunes itself in real time.

Caveat emptor! Most of the point-based lead scoring solutions purport to use machine learning but I have yet to find one that actually lives up to that claim. Most are building predictive analytics in a batch setting using data scientists (who also happen to use “machines”. I call them computers). If you’re truly looking for a machine-learning solution get educated and do some digging before you buy or upgrade.

Most of us may not care about machine learning, so regardless of where you are, there are steps you can take to implement more effective lead scoring. And these are not strategies that take a year to deploy; experience has proven that better outcomes can be realized in only a matter of weeks. So start off your new year right –with a leg up on your lead scoring.