Successful companies that use marketing automation solutions constantly evaluate their lead scoring models with a closed loop system of data collection, analysis and adjustment. Lead scoring is most typically evaluated by comparing actual results to predictiv. Are "hot" prospects leading to a consistently high number of closed deals? Is there a disconnect between what the scoring data tells you and what the sales reps actually see?
Creating a lead scoring team that includes both sales and marketing stakeholders works very well. Your sales managers and field reps can tell you which traits define their ideal prospect and which traits suggest a poor one. The marketing team, in turn, can turn this real-world feedback into a lead scoring model.
When your sales reps spot problems with the process, they need to let you know immediately. It's also important to review the data from your marketing automation, CRM, and sales force automation systems to look for patterns that show whether your current lead scoring model is delivering the goods.
There's no such thing as a "final" lead scoring model. In fact, the testing and tracking process should constantly identify new ways to refine your lead scoring model. If your sales pipeline data suggest that certain "hot" scoring criteria don't matter, for example, then your next step is to figure out which data points will be more useful for identifying hot prospects.