Common Mistakes in Scoring Models

Not applicable

I see a lot of scoring models in Marketo that seem to have been set up once, when the system was first implemented, and never revisited. Often times, these are not working as expected. A scoring model, like almost any operational program in Marketo, needs to be reviewed on a semi-regular basis.

In optimizing scoring, ideally you would have two stages of review:

  • Phase I: Review to discover if the scoring model you already have is working as you originally intended and adjust as needed.
  • Phase II: Review to discover if the scoring model designed is actually the best scoring model for your business and adjust as needed.

Since most of us are not yet doing Phase I, I want to start there and share some of the most common issues I see in scoring models. This is my personal list of “don’ts”:

  • Scoring on Email Opens: It’s an unreliable metric and is not always even a sign of engagement, as subject lines or from names may mislead uninterested individuals into opening the email. It can very easily inflate your behavior score unnecessarily.
  • Scoring on Tradeshow Attendance: Individuals attend tradeshows for so many varied reasons that it is not a reliable indication of buying intent in most cases. Instead we should focus on scoring individuals who show some meaningful engagement at the tradeshow.
  • Scoring Every Web Page Visit: This accumulates too quickly and tends to be overweighted when used. It is not always a sign of buying intent because many pages on your website might attract non-sales traffic (support pages, career pages, etc). Instead, you might want to score on specific key web pages or multiple web page visits.
  • Accidentally Double Scoring an Activity: Imagine you are giving out points for clicking a link on a webpage and for visiting a page. If I click a link on your home page to go to a resource page, I get 5 points for clicking the link to the resource page and another 5 points for landing on the page once it loads. This is basically double scoring someone for that same activity. I also see this happen a lot with form fills for event registrations.
  • Not Limiting Frequency: So many people leave the qualification rules on their scoring campaigns set to allow leads to run through every time. This causes scores to accumulate really quickly. Even something like Fills Out Form and Program Status Was Changed triggers can be set to only allow people to run through once an hour, which can help minimize double scoring due to errors like someone spam-clicking on a submit button that’s slow to load or an admin needing to delete and reupload attendee lists into the database.
  • Scoring Every Lead in the Database: Don’t waste database resources and your team’s time by scoring records that will never be prospects for your company. Exclude employees, competitors, partners, existing customers, etc. as appropriate for your business. Consider separate scoring models if you want to track behavior of other audiences.
  • Overweighting Demographic Scores: Marketo is all about reacting based on people’s engagement with your brand. If your scoring model treats demographics as equal to behavior, you may be missing the point. Obviously there are some business models that are exceptions but I think generally demographics work best when they are weighted to be about 1/3 of the total lead score, not 1/2.

Those are the main mistakes that I see. Feel free to add other suggestions in the comments!