Lead Scoring Bible

Tyron_Pretorius · ‎06-01-2024

If you are new to Marketo the prospect of building a lead scoring mechanism can be intimidating. There is alignment needed across the sales and marketing teams to define the model, which is never easy to get, involving a lot of meetings and communication, and then there is the implementation of the scoring mechanism in Marketo which can be equally challenging.

While getting alignment between the sales and marketing teams is not something I can help you with this post will give you some

examples of lead scoring models for demographics and firmographics as well as lead scoring models for activity. I will then help you define a marketing qualification mechanism (MQL) that will use the outputs from our lead scoring models to determine whether someone should be handed off to sales.

Not only do I outline some great scoring model examples I also link to the blog posts and videos that will show you how to implement these models in Marketo.

I reference Clearbit a lot in my examples below. If you have recently purchased Clearbit or you are interested in seeing how it works with Marketo then check out the Integrating Clearbit with Marketo post.

Demo & Firmo Lead Scoring Model 1

This lead scoring model is used to populate 2 fields:

Lead Quality
Lead Quality Detail

The Lead Quality field can have 1 of 4 values:

Top Match
Mid Match
Low Match
Unknown

Lead Scoring Model 1

Whether a lead falls into the Low-Top match buckets is determined by the values they have for the fields used in the lead scoring model. If the person does not have any values for the fields in the model then they are put in the Unknown bucket.

Feedback from the sales team on previous scoring models highlighted the need to not only let the sales team know how good a fit the person was but also why they were deemed to be a good fit. The reason a person is given a certain Lead Quality value is therefore recorded in the Lead Quality Detail field.

Fields Used in the Scoring Model

The first step in building any lead-scoring model is to use your customer profiles to determine what fields should be included in your lead scoring model. Some examples are given below.

Target account status: This is a boolean field populated by your sales team in Salesforce and it will be equal to true for anyone who is a member of a target account that your sales team is going after.
Clearbit Company Tech: This is a list of the technology platforms a company uses. For example, you could check this field to see if the person’s company is using any of your competitors’ software products so you know they will be a good fit for your company.
Clearbit Employee Number: This can be used as a gauge to determine the size of a company with the assumption that the higher the number of employees the bigger the potential opportunity that could be generated.
Clearbit Industry: This is an easy addition to any scoring model since you can only sell to people in certain industries so you want to score these people higher.
Job Title: This can be populated from Clearbit or it can be populated from forms on your site or sales team input.
Clearbit Alexa Ranking: This field gives an indication of how prominent the company is on the web with the assumption that the more prominent they are the bigger and the more established the company.
Clearbit Company Description: This field is a gem and allows you to score on certain keywords related to your product or service that are present in this description.
Clearbit Company Tags: There is some overlap here with the Clearbit Industry field but the tags can have more granular or specific values that you are interested in.

You can see all the available values for the Clearbit fields mentioned above from either the Clearbit website or their attribute values sheet.

Building Fields into Scoring Logic

Once you have determined what fields you are interested in scoring on the next part is to determine how you should score the values in these fields and the different combinations someone can have for these different field values.

The “Lead Quality Scoring” sheet shows how the different combinations of these fields and their values will determine someone’s Lead Quality and Lead Quality Detail.

There are 2 additional tabs “Company Description Keywords” and “Job Title Keywords” which contain the Lead Quality and Lead Quality Detail values assigned to each keyword match for the Clearbit Company Description and Job Title fields.

Now that you have your scoring model outlined you can check out the ICP Matching blog post to see how to implement this lead scoring model in Marketo.

Demo & Firmo Lead Scoring Model 2

Lead Scoring Model 2

The output of this model is 1 single field called “Quality Score Tier” which can have four different values: A, B, C, or D.

The combination of the “Demographic Score Tier” and the “Firmographic Score Tier” determines the “Quality Score Tier” value that a person will have according to the matrix in the image above. For example, if someone has Demographic Score Tier = Tier 2 and Firmographic Score Tier = Tier 3 then we can see from the diagram above that they will have a Quality Score Tier = C.

The “Demographic Score Tier” and the “Firmographic Score Tier” fields are calculated from the sum of all the demographic scores and the sum of all the firmographic scores divided by the maximum possible demographic or firmographic scores.

Demographic Score Fields

The Demo Score Control Panel sheet determines the scores for each of the 5 demographic score fields in the model depending on what values the corresponding fields have:

Demo Score – Email
Demo Score – Seniority
Demo Score – Title
Demo Score – Phone
Demo Score – LinkedIn

Scoring logic for demographic fields

All 5 of these scores are summed up and divided by the maximum possible demographic score (95 in this case) to get a decimal value ranging from 0 to 1. Then the “Demo Total Low Threshold” and “Demo Total High Threshold” values are used to determine the “Demographic Score Tier” for the lead. Taking the thresholds from the sheet of 0.3 and 0.6 for the low and high thresholds respectively if a lead has a decimal value:

> 0.6 then Demographic Score Tier = Tier 1
>=0.3 & < 0.6 then Demographic Score Tier = Tier 2
<0.3 then Demographic Score Tier = Tier 3
= 0 then Demographic Score Tier = Tier 4

Firmographic Score Fields

The Firmo Score Control Panel sheet determines the scores for each of the 5 firmographic score fields in the model depending on what values the corresponding fields have:

Firmo Score – Website
Firmo Score – Revenue (EAR = Estimated annual revenue)
Firmo Score – Alexa
Firmo Score – Industry
Firmo Score – Country

Scoring logic for firmographic fields

All 5 of these scores are summed up and divided by the maximum possible firmographic score (100 in this case) to get a decimal value ranging from 0 to 1. Then the “Firmo Total Low Threshold” and “Firmo Total High Threshold” values are used to determine the “Firmographic Score Tier” for the lead. Taking the thresholds from the sheet of 0.3 and 0.7 for the low and high thresholds respectively if a lead has a decimal value:

> 0.7 then Firmographic Score Tier = Tier 1
>=0.3 & < 0.7 then Firmographic Score Tier = Tier 2
<0.3 then Firmographic Score Tier = Tier 3
= 0 then Firmographic Score Tier = Tier 4

Input Demographic & Firmographic Fields

You can see all the available values for the Clearbit fields mentioned below from either the Clearbit website or their attribute values sheet.

Demographic Fields

Email Address: Business emails are given a higher score than freemails
Seniority (Clearbit field): Depending on your ICP you could score people from the C-suite higher than managers or individual contributors
Job Title: As well as being another indicator of a person’s seniority at a company, you can also score on keywords in job titles that match your ICP e.g. +10 points if “developer” in job title
Phone Number: If they have provided a phone number this shows more intent for them to get in touch with your sales team so they should be scored higher
LinkedIn Profile URL (Clearbit field): If you have a LinkedIn profile URL for a lead your sales team will be able to find out a lot more information about this person and this also offers another avenue for outreach so people with this field should be scored higher

Firmographic Fields

Website: Certain website domains such as “.com” indicate better quality than domains such as “.net”
Estimated Annual Revenue (Clearbit Field): You can use the revenue generated by a company to assess whether they fit into the company size you are targeting e.g. startup versus enterprise
Global Alexa Ranking: This field gives an indication of how prominent the company is on the web with the assumption that the more prominent they are the bigger and the more established the company.
Industry (Clearbit field): This is an easy one to include since you only want to score leads highly if their company is in an industry that could use your product or service
Country: You may only do business in certain regions of the world or you may know that leads from certain countries e.g. US, are likely to be better quality than leads from other countries e.g. Nigeria.

Testing your Lead Scoring Model

One way you can fine-tune how you are scoring the different demographic and firmographic fields is to import some leads from your Marketo database into the Quality Scoring Calculator sheet and see what scores they are given.

Then you can change the parameters in the Demo Score Control Panel and Firmo Score Control Panel sheets and automatically see these changes applied to the scores of the sample leads.

To import and score sample leads:

Open the Quality Scoring Calculator sheet
Import the CSV containing values for all the demographic and firmographic fields that are being scored as a new tab in the workbook
Copy the name of this tab and paste it in the cell beneath the “Sheet Being Tested” header in the “Quality Score Calculations” tab
The uploaded people information from the named sheet will be auto-populated and the field scores as well as the demographic, firmographic, and quality scores will be calculated

Once you have fine-tuned the parameters in your lead scoring model you can then check out the Demo & Firmo Grading blog post to see how to implement this model in Marketo.

Demo & Firmo Lead Scoring Model 3

AI tools are transforming the way we work as marketing operations professionals and lead scoring is yet another avenue where AI tools can assist us.

The most important and hardest part of lead scoring with ChatGPT is developing the right prompt. As shown in the Integrating ChatGPT with Marketo blog post, implementing lead scoring with ChatGPT is pretty straightforward using Zapier or Marketo webhooks.

I found the best way to test out different prompts is by importing a list of leads, consisting of leads you know are good and bad quality, into a Google Sheet and then using the GPT for Sheets and Docs app to score the leads with ChatGPT.

You will need to get an API key from your ChatGPT account to set up the GPT for Sheet and Docs app.

Lead scoring with ChatGPT

You can then use Google Sheets formulas to determine how well a prompt performs when comparing the output GPT lead score with your human determination of whether the lead is a good fit (“Human Decision” column) and any existing lead score fields that you use (“Lead Quality” column).

I highly recommend checking out the walkthrough video in the Integrating ChatGPT with Marketo blog post to see how to use the GPT formula and how to iteratively test and compare the performance of different prompts.

Comparing GPT scoring performance to existing lead scoring

Lead Scoring for Activity

At your company, I am guessing there are a number of high-intent activities like filling out a contact sales form that you MQL for immediately. However, what happens when a lead does lots of smaller activities (e.g. viewing 20 web pages and downloading an ebook) which when considered together mean that they might be ready to talk to sales?

This is where a field like the “Behavior Score – 7 Day” field is helpful. Every activity that a lead can carry out is given a score and all the scores for the lead’s activities over the past 7 days are summed up and stored in the “Behavior Score – 7 Day” field.

This way a lead’s intent over the past week can be assessed using a single field and different “Behavior Score – 7 Day” thresholds can be applied to leads of different quality. For example, your top-quality leads might need only 10 points to be handed off to sales whereas the low-quality leads need 20 points to be handed off to sales.

Also to give sales more context on the past activities that a lead has done it is useful to have a “Behavior Score – 7 Day History” field which shows the “Behavior Score – 7 day” value and the activities that contributed to this value each time the lead went through this scoring model (see example below).

{
   "timestamp":"2021-05-18 07":"01":09,
   "Portal [Added Numbers to Cart]":+2,
   "Portal [Checked Portability]":+2,
   "Behavior Score - 7 day":4
},
{
   "timestamp":"2021-05-21 07":"01":06,
   "Portal [Checked Portability]":+2,
   "Behavior Score - 7 day":2
}

Marketo Activity Lead Scoring Model Definition

The first step of building an activity scoring model like this is to collect all the activities that you can track in Marketo in a Google sheet and then decide on a point value for each activity. For an example of how you can do this, you can take a look at the “Lead Scoring for Activity” sheet.

In columns E-G the activities are grouped by their “Type” e.g. all page view activities are grouped together so that it makes it easier to compare the score for activities with the same type. For example, this makes it easy to see if viewing a competitor page should be worth 3 points when viewing 10+ generic pages is worth 2 points.

Grouping activities by type

In columns A-C the activities are sorted in terms of their score so this makes it easier to compare the score for activities across the different types. For example, this makes it easy to see if purchasing more than 10 numbers should be equal to the same number of points as a contact sales form fill.

Sorting activities by score

Next, you need to decide the point threshold for each lead quality that will cause these leads to be handed off to sales. For example:

A quality needs a score of 10 to be handed off to sales
B quality needs a score of 11 to be handed off to sales
C quality needs a score of 12 to be handed off to sales
D quality needs a score of 15 to be handed off to sales

Once you have all this stuff figured out head on over to the Activity Tracking blog post to see how to implement this lead scoring model for activities.

MQL Model Outline

Once leads have been scored using your model of choice it is now time to combine these lead scores with activity classifications, which will then determine if someone becomes an MQL and gets handed off to sales.

As we have seen with the activity lead scoring model there will be lots of smaller activities that won’t warrant a lead becoming an MQL, however, if lots of these smaller activities are done together then this could push the lead over the points threshold for their lead score and lead to them MQLing (see the points thresholds in the section above).

In addition to the activity lead scoring model, we can also have activities that trigger MQLs immediately based on the lead score.

For example in the MQL Activities sheet, activities are classified in terms of their “Priority” where:

Critical activities should trigger an MQL no matter the lead score
Important activities should trigger an MQL for leads whose lead score is A-C or Low-Top match
Normal activities should trigger an MQL for leads whose lead score is A-B or Mid-Top match

Categorizing activities by priority

When an MQL gets handed off to the sales team, the sales team members can use the MQL Source and the MQL Source Detail fields to get context on why the lead became an MQL.

Take a look at the MQL Mechanism post to see how to implement this MQL model in Marketo.

What’s After Lead Scoring?

Now that you have covered one of the fundamentals of Marketo you might be interested in mastering other core concepts such as: