Predictive Outcome Scoring Bridges Marketing and Sales
Lead scoring originated as a way to identify Sales Qualified Leads or SQLs, in customer relationship management (CRM) software platforms. Thanks to open API integration, B2B companies now have access to resources that just a few years ago, only enterprise firms had access to. Small and medium size firms can now connect and synchronize almost any database set far beyond the demographics and actions that lead scoring originally provided. I think the terminology – lead scoring – is a bit out of date though. Predictive Outcome Scoring or Engagement Scoring seems a more accurate description in today’s technology-driven marketing activities because you can set parameters to track and measure a much broader range of data, actions and behaviors.
Identifying when top of the funnel leads are adequately engaged and qualified to pass over to sales is a hot topic for many marketers. Engagement scoring applies values to behaviors and actions to helping marketers identify when a lead has the highest probability of being ready to engage with your sales team. I decided to put my Excel-erator hat on and have a go at creating a tool for marketers that may help you score the visitor behaviors that are most relevant to your individual needs. Instructions follow and are broken down into four steps:
- Identify Objectives and Parameters
- Behavior Values
- Test Your Values
- Determine Your Pass Over Score
I hope this scoring tool helps guide you on assigning values to behaviors, actions and demographics as you create your scoring and grading strategy. I want to emphasize, this tool is intended as a guideline to help you visualize and understand how your scoring values rank and how rules pass over scores between platforms. Please keep in mind, every company is situational with unique requirements and objectives as well as different software platform(s), integrations and configurations. If you choose to download the tool, I hope it helps you. I welcome your feedback, comments and thoughts.
Before we dive into creating a scoring metric, let’s review a few basics.
Scoring Parameters
Data collection for any scoring metric can come from several technology sources:
- Customer relationship management system
- Lead management system
- Internal data
- Social channels
- Email automation system
Data rules can identify and measure:
- Behaviors and engagement
- Demographics
- Actions (external and internal)
This is a good time to do a review of your software systems and understand at a high level how information is being passed from one platform to another. If you have a lead management system or email automation system connected to your CRM, you should be able to identify, track and measure certain levels of engagement from first touch to customer renewal and all points in between. Little happens if a visitor to your website, which is a potential customer, is not engaged. You can use scoring metrics as a signal(s) to pass information to your team, track engagement through the marketing and sales funnel and analytic signal to help in identifying which behaviors, content, campaigns and social activity best contribute to your ROI.
Once you have reviewed your software field values and mapping rules, you are ready to build your engagement scoring strategy. The objective of the scoring tool is to assign values, see how they score collectively in different scenarios, create “what if” scenarios to help you identify more accurately the Total Score pass rule you will use to qualify leads as MQL that will automatically send to your CRM.
Defining Variables
Step One: Identify Objectives and Parameters
Make a list of all the ways a potential lead could engage with your brand such as:
- Website/blog pages visited or specific pages visited
- Number of returning sessions
- Complete an action such as a trial or contact us form
- Download an offer
- Register for a webinar
- Attend a webinar
- Number of email opens
- Number of clicks on emails
- Facebook content clicks
- Twitter content clicks
- LinkedIn content clicks
- Google+ content clicks
- Subscribe to newsletter
- Register for an event
- Attend an event
- Participate in a campaign
- Attend a tradeshow
If demographics are a valuable tool to help you identify when a lead becomes an MQL and your software limits you to a rule such as “filled out a form” but not a rule that identifies how many forms filled out, consider this work-around. If the platform managing your forms has progressive profiling, you may be able to use the “has this value” rule to identify a demographic. For example, if country is the 2nd field in your progressive profile list, then adding the country in a form fill would be the 3rd time a visitor has filled out a form. That may allow you to identify a multiple form field in a slightly creative way. Note: It may not be completely accurate, especially if there are other ways the country field can be added to leads, but it is a good alternate if you identify that multiple downloads have engagement value.
Step Two: Behavior Values
Analyze how valuable these collective behaviors and demographics are in accessing when a lead is warm enough and has shown enough interest for the lead to be passed over to sales. In our scoring tool, we have created a set of 3 models for you to create what if scenarios. The data you enter in Scoring Test Version A is replicated into Version B and C. The colored areas are formula fields. As you add numbers in the white cell areas, the colored section will re-calculate based on your entries. Note: You do not have to add the percentage sign; entering the number 50 will appear as 50%.
I used percentages to help determine which behaviors you decide should be the most weighted. For example, requesting an action such as a contact form is probably important 100% of the time. Downloading an eBook is sometimes a first touch point. Therefore, you may want to consider a free content offer or eBook offer may only be important 30% of the time. Signing up for a webinar may have an importance of 60% and attending a webinar may be important 80% of the time. Once you assign the importance values, you can make slight adjustments in Version B and C to compare.
Step Three: Test Your Values
Determining your score values and which values work best is a critical pre-planning criterion. Once your system begins calculating scores, many systems only enter the values one time and do not re-calculate pre-existing scores. You can change the total score that will trigger a notification to your CRM, but avoid changing behavior values once the system is implemented. If you decide to change your values, past leads may have one set of scores, and future leads will have another set which may compromise your ability to use scoring for analysis.
If you are planning on using engagement scoring for determining when a lead is marketing qualified or to identify engagement levels in customer retention, or a lead scoring metric for moving SQL leads through the sales funnel, your scoring model between departments needs to be consistent. In other words, you do not want one set of metrics to be based on a score of 1 – 5 and a different model to be based on a score of 1 – 10. Also, be mindful of not spreading your scoring range too far. Your lowest value should not be more than 30% to 50% of your highest value. No single behavior should be equal to your total score. For example, if you want contacts that fill out a contact us form to be sent to your CRM regardless of scoring, consider adding an exception rule. Add a hidden field to your form with a specific value, then assign the rule to your CRM to always pass this field value.
Our scoring tool allows you to simulate behaviors so that you can pre-determine how leads are scored based on different criteria by adjusting your top score value. Add a 1 in the cell to the immediate right of each behavior. Your pre-calculated scoring variable will automatically appear under the Score Simulation column. You can also change the name of the behaviors to suit your needs. This will help you determine the collective score for qualifying a lead as MQL-ready.
Additionally, if you want to envision variations to your weighted values, you can set the Top Values to the same value, e.g. 10 and adjust the percentage of weighted value in Versions B and C to compare the difference from Version A. To revert to the same percentages as in the Version A column, simply add this formula into the formula bar: “ =$D6 ” and copy the cell into any of the percentage cells you changed:
So far, you have identified your variables and should have a better idea of how your scoring model is going to be calculated. Keep in mind, these are only guides. Software platforms vary on what types of behavior data, and how data is queried. The scoring tool is also a great resource in collaborating with the technical teams on your various platforms of what you want.
Your next step is to identify how valuable repeat activities are in defining a predictable model that triggers when a top of the funnel lead is engaged enough with your brand to become a MQL lead. Keep in mind, not every behavior you have identified may be a logical query for your platform. For example, in one lead management system, I can identify specific pages a visitor visits but I cannot set a rule for how many pages they visit.
Applying Variables to Scoring Parameters
In our Engagement Scoring Tool, open up Tab 2: Scoring Parameters.
Step Four: Determine Your Pass Over Score
In most software lead scoring rules, once a visitor has accomplished a behavior, e.g. downloaded an eBook or an offer, the scoring is only applied once. For example if:
- Visitor A downloads one eBook, subscribes to your newsletter and registers for a webinar and
- Visitor B downloads 3 eBooks, subscribes to your newsletter and registers for 4 webinars
Visitor A and Visitor B will both have the same score unless you add additional parameters to your scoring rules. The names and values you created in the 1st tab have been populated into Tab 2 to make the work you have done so far easier. The names and percentage values you created in the previous tab are highlighted in gray. Simply add the additional behaviors and values. There are also blank rows further down. Once you add a behavior and value, the blank fields were also re-formulated. Note: Refer to your platform(s) on their capability on setting additional queries.
Once you add and/or adjust the remaining percentages, you can create some scenarios that will allow you a better understanding of how and when leads will pass the MQL criteria. I prefer using a 1 – 10 scoring rule as it allows you a little more control of the “gray area.” In the following example, if you were to use a 1 – 5 score, the middle scenario would be to hold for more nurturing. The same criteria using a 30 pass score passes as a MQL but stays as a nurturing lead when a 35 score is applied. The differences are subtle, yet as you apply additional behaviors and rules as marketing changes often demand, the 1 – 10 scoring offers a slight advantage in the control in your outcomes.
Setting up your lead scoring method is not a simple exercise. The rules between your marketing platforms need to be fully sync’d and collaboration between team members managing each platform must be collaborative on all changes as one set of rules on one platform can affect rules on the platforms you are synchronizing.
I hope our scoring tool and this post offers some clarity on the topic of engagement and predictive outcome scoring. I also hope it helps you to avoid some common pitfalls while building an effective scoring tool for your inbound marketing strategy. I welcome your thoughts, questions and comments. And of course, if you need help with this, let us know:)
