Here’s a quick look at my findings from optimizing our lead scoring model over the past few months. Comment below to add your own thoughts and feedback.
4 Actionable Tips:
1. Translating the scoring model for sales
I came across this while trying to get feedback from sales as to how to improve the scoring system. The +10 and -5 didn’t really resonate with the group, so for every action I added a temperature label to communicate how “hot” or “cold” a lead was based on the action.
This way it is easy to speak the same language and gain consensus as to what actions deserve each score.
2. Clump Analysis
There is probably a more scientific/statistical term for this – but the idea here to export all of your leads into excel with a few columns displayed. These columns would have to be your scoring fields, and then include other interesting fields like lead status. If your database is very large, you may want to export a segment (for example, leads in the last 6 months).
Using the Countif function in excel you can easily come up with a frequency distribution. Or you can use a method like this: http://www.excel-easy.com/examples/frequency-distribution.html
Invariably you will find large clumps of leads sitting at certain scores, which can provide insight into your entire model. However, a quick win will be to evaluate this against your MQL threshold score. For example, if sales have the bandwidth to review more leads - or if they lack time to review the current leads assigned to them, you know how many notches to kick the score up or down.
3. Scoring Footprint at the campaign level
If you have not conducted a scoring footprint audit, it is very helpful and instructions are here: https://nation.marketo.com/blogs/marketowhisperer/2015/12/09/4-simple-ways-to-evaluate-if-your-scoring-model-works . This is the next level down, to see which scoring campaigns have influenced your leads in a given time period.
For each scoring smart campaign, check the members by week tab in the summary view. First, check each one to make sure that members are running through the program. If there are no members in the program, either the campaign was setup incorrectly, or there is a process problem downstream.
4. Evaluate 10 most recent closed/won deals, 10 most recent disqualified/junk leads
The recommended method to develop and evaluate scoring is to analyze commonalities among a large number of closed deals and disqualified leads. However, analyzing the most recent won deals and disqualified leads offers an excellent barometer to how accurate your scoring model is. You can use the opportunity analyzer, but for me creating an excel sheet of the company size, industry, title, and then the key activities that were performed before close is even more insightful. To start attributing negative scores to bad leads, find patterns among the most recent disqualified leads. Company too small? Student or intern title? Bad industry fit? Start assigning negative points for each of these and try to add more on a regular basis.
Hope some of these tactics help – would love to hear your favorite lead scoring optimization tips!