How to Build a Lead Scoring Model

Automation is the only way to qualify inbound leads at scale. And a lead scoring model is the crux of an automated qualification flow. So how do you build a model that works for the modern B2B buying journey? Read this article to find out. 

What is a lead scoring model?

Lead scoring models measure various lead attributes and behaviors—often hundreds of individual data points—and weigh them to calculate an overall score. The score determines whether a lead meets enough criteria to pass your qualification threshold. 

‍

Factors used in building your lead scoring model can include:

‍

• Individual demographics 
• Company firmographics
• Behavioral data
• Intent data
• Engagement with specific marketing campaigns

Why is a lead scoring model important?

According to Adobe, 96% of website visitors aren’t ready to buy. And data from Invesp shows that among those that end up converting, only 20% will close. So most of the people you engage won’t buy. 

‍

To avoid wasting time on dead ends, you need an automated way to surface high-intent leads. That’s where a lead scoring model comes in. By scoring and identifying high-value leads among your hundreds and thousands of conversions, you can prioritize sales outreach to those most likely to close. 

‍

Lead scoring provides a number of additional benefits as well: 

‍

• Identify low-intent but high-value leads for nurturing
• Identify low-value but highly engaged leads for unqualified lead followup
• Identify current customers who show heightened interest in certain products or features, opening opportunities for cross-sell and upsell
• Find outliers based on individual attributes or behaviors
• Tailor each lead’s buying experience to their level of interest, improving overall satisfaction

How to build a lead scoring model

So how do you actually build a lead scoring model that works for modern B2B? For starters, you need a clear lead qualification strategy, developed ideal customer profiles (ICPs), and a map of their buying intent and behaviors. 

1. Choose your lead scoring model

Different lead scoring models work better for different organizations. As such, the first step to building a lead scoring model is to determine which one you need. Here are some of the more common ones in B2B:

‍

• Purchase intent model. This lead model prioritizes buying intent, measured through first- and third-party intent data, to find those leads most interested in your product, regardless of whether they align with your ICP.
• ICP-based model. This model weighs ICP alignment more heavily than intent or engagement data. 
• Engagement model. This model looks primarily at engagement with your content or brand, prioritizing those leads most interested in your company. 
• Negative model. Rather than surface high-value prospects, a negative model focuses solely on excluding low-value prospects. These can be indicated by email unsubscribes, career page visits, spam submissions, internal IP addresses, and associations with competitors. 

‍

Regardless of the model you choose, it’s important to identify exactly what you want to accomplish and where your priorities lie. This will provide much-needed clarity as you build out your model. 

2. Map out your customer journeys

Once you have a strategy in place, you need to start to map out your ICP buyer journeys. Since B2B customer journeys are non-linear, this is a more complex exercise than simply drawing a funnel. 

‍

The first step is to start with the five main stages of the customer acquisition lifecycle: awareness, engagement, consideration, decision, and renewal. While your lead scoring model will focus primarily on the first three stages, it’s important to have an idea of what a quality purchase decision or post-purchase success looks like. That way, you can further fine-tune your model around leading indicators of a high LTV. 

‍

Once you have the basic customer journey in place, it’s time to muddy the waters. Start modifying your journey map with the more complex behaviors leads may engage in: omnichannel touchpoints, double backs, behavioral triggers, decision-making units, and dynamic engagements. Only then will you have a clear idea of what the actual journey looks like. 

3. Assign points to key criteria

Once you consolidate those key attributes, behaviors, and intents into a coherent customer journey, start assigning points to each of them. This is where the actual lead scoring model starts taking shape. 

‍

Every lead scoring model includes the various data mentioned above. However, your specific model will weigh the factors you prioritized in Step 1 above others. For instance, an intent-based model will weigh high-intent data (e.g. demo page visits) more heavily over firmographic data (e.g. annual revenue). 

4. Set up automations around lead scores

Once you have your lead scoring model in place, you need to set up automations to take action based on specific lead scores. These can include:

‍

• Qualification workflows. Route qualified leads directly to the appropriate salesperson and automate scheduling to avoid meeting dropoff.  
• Nurture workflows. Add partially qualified leads to a nurture sequence to educate them and drive further interest and engagement. 
• Outlier workflows. For leads that may not be fully qualified but feature some indicators of a potential deal (i.e. high engagement but no ICP alignment), set up automated workflows to manually determine whether they should go to sales. 
• Non-qualified workflows. For leads that aren’t qualified but can result in valuable conversations and insights, route to other internal contacts—marketing, product, or leadership—to initiate those conversations. 

5. Continually monitor and adjust

As you convert, qualify, and close more leads, you’ll naturally learn more about the market. This will require you to make adjustments to your lead scoring model. This could involve simply adjusting the weight assigned to each criterion, or adjusting the necessary threshold for qualification. 

‍

These changes should happen occasionally and only on the margins. If you find yourself making regular, significant adjustments to your model, you probably made some errors in earlier stages.
Go back and modify your strategy rather than keep trying to fix what’s already broken. 

Final thoughts on building a lead scoring model

The whole point of building a lead scoring model is to enable seamless automation and scalability in your inbound flows. While some regular adjustment is necessary, it shouldn’t require constant manual intervention.

‍

One of the reasons lead scoring models fail is because broken integrations among point solutions result in lost or bad data, as well as workflow misfires. The best way to avoid this problem is to implement a unified GTM platform that qualifies, routes, schedules, and nurtures all in one place. 

‍

See Default’s unified GTM workflows in action here. 

‍