When is the last time you audited your lead scoring model?

If there are no complaints from sales that means it’s working, right?

The only way to determine how well your lead scoring is predicting leads is by looking at the success of the leads that have qualified. How many of those leads went on to be opportunities? How many went on to be rejected? Often times as marketers, we are only looking at the top half of the funnel - getting leads to sales - but it’s time we think about the bigger picture.

Assuming you have a few basic data points such as MQL date, Nurturing/Disqualified Reasons, and Lead Source, here are a few things to look at to evaluate how well your lead scoring model is predicting leads.

When pulling these reports, you’ll want to evaluate a time period that is long enough to account for your sales cycle, allowing time for your MQLs to become opportunities.

Status: MQL > Recycled

Pull a report of leads that became an MQL (Filter: MQL date is not equal to blank) and have gone to “Recycled” or “Nurture” status (Filter: Lead status = Recycle/Nurture).

• Report 1: Summarize the data to look at those leads’ Lead Source/Lead Source Detail and Most Recent Lead Source/Most Recent Lead Source Detail.
  
  How can this help evaluate scoring? If a majority of the leads have the same lead source or lead source detail, it may be an indication that your behavior scoring is too high for a certain activity. For instance, if many of the leads have become MQLs from content downloads and are then getting recycled, that you could tell you that behavior scoring for content downloads is too high and that leads need more activity before they are ready to be passed to sales.
• Report 2: Summarize the data to look at Recycled or Nurture reason.
  
  How can this help evaluate scoring? Look for trends. If these leads have a recycled reason in common such as “Bad timing,” that could indicate that your behavior score threshold is too low, and that leads need more activity before moving to sales.

Status: MQL > Rejected

Leads that have moved from MQL status (Filter: MQL date is not equal to blank) to a “Rejected” or “Unqualified” status (Filter: Lead status = Rejected/Unqualified).

• Report 3: Summarize that data to look at those leads’ Lead Source/Most Recent Lead Source.
  
  How can this help evaluate scoring? This could help flag a lead source that is bringing in sub-par leads. For instance, if many leads are coming from Content Syndication and then getting rejected by sales, behavior scoring could be adjusted so those leads aren’t scored for their Content Syndication activity.
• Report 4: Summarize that data to look at Rejected or Unqualified reason.
  
  How can this help evaluate scoring? If leads have a rejected reason in common such as “Not a fit” it could be an indicator that your demographic scoring is qualifying leads that aren’t the right business fit.

Status MQL > Opportunity

Leads that became an MQL (Filter: MQL date is not equal to blank) and are now part of an opportunity (Filter: Opportunity Name is not equal to blank).

• Report 5: Summarize leads by Lead Source and/or Lead Source Detail.
  
  How can this help evaluate scoring? If you’ve seen a common theme in where these successful leads are coming from, that can tell us which channels result in the most opportunities. You may want to weight the behavior score for those channels higher than others.

Running these types of reports should provide some insight into what is happening to leads after they are passed to sales, which can help you make informed updates to your lead scoring model. Lead scoring shouldn’t be a “set it and forget it” project. You can repeat this exercise quarterly to see if your MQL to Opportunity ratio is improving and see where there is room for improvement.

Pro tip: before making changes to lead scoring, build your scoring model in the Lead Scoring MQL Simulator, so you can simulate your changes before implementing them in production. Happy scoring!