Health score logic: 8 rules for weights and criteria

Health score logic in customer success tools depends on two distinct mechanisms: how individual scores combine into an overall metric and how specific conditions trigger score updates. Understanding these mechanics is essential for configuring accurate account health assessments.

In Vitally, the calculation method is explicitly defined. According to Vitally, an account's overall health is then calculated as a weighted average of each health score. This means that if you create multiple distinct health scores for different aspects of customer engagement, the final health indicator is not a simple sum or a single highest score. Instead, it is a composite figure derived from the weighted average of those individual scores. When configuring this in Vitally, you must ensure that the weights assigned to each component score reflect the relative importance of that metric to your overall churn risk model. The platform handles the arithmetic of this weighted average once the individual scores and their respective weights are established.

If you are setting up a new health score or auditing an existing one, check the documentation for your specific platform. For Vitally, confirm the weighted average calculation. For HubSpot, confirm the AND logic for criteria. This ensures that your health scores reflect the intended logic and provide accurate insights into account health. The difference between a weighted average and strict AND logic is significant. A weighted average allows for partial contributions from multiple metrics, while AND logic requires all conditions to be met for any change to occur. Understanding this difference is the first step in configuring health score logic correctly.

• Vitally calculates an account's overall health as a weighted average of each health score, as stated in Vitally documentation.
• HubSpot uses AND logic for score criteria, meaning the score updates only if all criteria are met, according to HubSpot.
• Gainsight restricts scorecard criteria by excluding Date, Calculated, and Look-Up fields, as noted in Gainsight guides.
• HubSpot automatically creates Health score and Health status properties when a health score is turned on, according to HubSpot.
• Gainsight Unified fields inherit properties like Data Type and Decimal Points from Master fields, reported by Gainsight.

The following table maps the eight specific configuration rules to the platforms that define them. Each row identifies the publisher and the exact logic or limitation documented in their respective guides.

Audit your current health score configuration to verify that weighted averages are applied correctly and that criteria logic matches your intended segmentation.

In the Vitally Help Center documentation, an account's overall health is calculated as a weighted average of each health score, according to Vitally. This calculation method combines distinct score components into a single aggregate value for the account. The platform supports multiple health scores at the Account or Organization level, allowing users to create separate scores such as one for "Product usage" and another for "NPS," as described in the Vitally documentation. This structure enables teams to track different dimensions of customer health independently before they are combined.

The interface presents these scores with a default breakdown view, which displays the individual health scores alongside a heat map, according to Vitally. The heat map is located on the bottom right of the view and shows a customer's health each day. The color intensity in this map indicates activity levels, with darker colors representing days when the account was more active. This visual representation helps users identify daily engagement patterns related to the calculated health scores.

When configuring these scores, ensure that the weights assigned to each component align with your strategic priorities for customer success. The weighted average logic requires that each score contributes to the final overall health metric proportionally to its assigned weight. The ability to maintain multiple score types, such as product usage and NPS, allows for a more granular assessment of account health than a single composite score would provide. Review the breakdown view regularly to monitor how individual score components contribute to the overall health calculation.

For a broader perspective on setting up these metrics, see Health score setup: 8 rules for weights and validity.

When setting up health score criteria, the logic governing how conditions are evaluated varies by platform. In HubSpot, the "Create a health score in the customer success workspace" feature uses specific logic for multiple criteria. According to HubSpot, this option follows AND logic, which means the score will update only if all criteria are met. This strict requirement ensures that a single unmet condition prevents the score from changing, forcing a holistic view of the account's status rather than a partial one. When auditing your configuration, verify that every condition you intend to trigger a score update is explicitly included in this AND chain. If you rely on OR logic elsewhere, you must map that difference carefully, as HubSpot’s documented behavior for this specific workspace feature is strictly conjunctive.

Field inheritance presents a different configuration challenge, particularly in systems with layered data structures. Gainsight’s "Rules Engine Horizon Experience Overview" documentation addresses how data properties flow between field types. As reported by Gainsight, unified fields inherit the properties of Master fields, such as Data Type and Decimal Points. This inheritance rule means that when you configure a Master field, the downstream Unified fields automatically adopt those specific attributes. You do not need to manually re-enter the data type or decimal precision for each Unified field; the system handles this propagation. However, this behavior is specific to Gainsight’s rules engine architecture. Do not assume that other platforms, such as HubSpot or Vitally, apply the same inheritance model to their respective field types. If you are migrating logic from one tool to another, verify whether the target platform supports automatic property inheritance or requires manual configuration for each field instance.

Beyond inheritance, you must account for limitations on which field types can serve as scorecard criteria. Gainsight’s "Create Scorecards" guide outlines specific restrictions for administrators. According to Gainsight, the Date, Calculated, and Look-Up fields cannot be selected as part of scorecard criteria. This limitation means you cannot directly use these three field types to define the conditions for a health score. If your current logic relies on a calculated field or a date-based threshold, you must refactor your approach. You may need to create a different type of field that is eligible for selection, or adjust your criteria to use fields that are not restricted by this rule. This constraint is critical for audit purposes: if your team’s health score model depends on calculated metrics, you must ensure those metrics are stored in a field type that Gainsight permits for scorecard criteria.

In the "Create a health score in the customer success workspace" guide, HubSpot describes a specific approach to segment targeting. The team applies the score to a segment of existing customers, then configures score criteria using both event groups and property groups HubSpot. This method allows you to focus your monitoring efforts on a specific subset of your customer base rather than attempting to score every single record in your system. For instance, the guide notes that Sprocket wants to monitor a defined group of existing customer companies rather than score every record in the account HubSpot. By defining this target segment first, you ensure that the subsequent criteria configuration is relevant to the specific customers you intend to track.

When you turn on the health score in HubSpot, the system handles the creation of necessary data fields automatically. Specifically, the properties Health score and Health status will be automatically created in your account HubSpot. This automatic property creation step simplifies the setup process because you do not need to manually define these core fields before activating the score.

To implement this in your own setup, identify the specific segment of existing customers you wish to monitor. Configure your score criteria to include both event groups and property groups that are relevant to this segment. Once your criteria are defined, activate the health score to trigger the automatic creation of the Health score and Health status properties. Verify that these new properties appear in your account and that they are populating correctly for the targeted segment. This ensures your health score logic is applied only to the intended customer group, maintaining data accuracy and relevance.

Document each score using columns for Platform, Score, Criteria Fields, Criteria Logic, and Weight. Fill in the details for each score you manage to ensure consistency in logic and weighting.

When populating the "Criteria Fields" column, note that Gainsight requires specific selection steps for certain data sources. For example, when scheduling rules, you select the S3 dataset from the Task drop-down menu, as described in the Gainsight documentation.

• Onboarding tools: compare 4 customer project plans
• NPS survey configuration: 5 checks before launch
• Free trial setup: 8 items and a 730-day Stripe cap

How is an account's overall health calculated in Vitally?

In Vitally, an account's overall health is calculated as a weighted average of each health score, according to Vitally. This method aggregates individual score components into a single overall metric for the account.

What logic does HubSpot use when multiple criteria are set for a score?

HubSpot follows AND logic for health score criteria, meaning the score will update only if all criteria are met, as reported by HubSpot. This strict matching requires every configured condition to be satisfied before the score changes.

Which field types cannot be selected as part of scorecard criteria in Gainsight?

In Gainsight, Date, Calculated, and Look-Up fields cannot be selected as part of scorecard criteria, according to Gainsight. These specific field types are excluded from the available options when configuring scorecard conditions.

What properties are automatically created when you turn on a health score in HubSpot?

When you turn on the health score in HubSpot, the properties Health score and Health status will be automatically created in your account, as stated by HubSpot. These properties appear in your account without manual creation steps.

How can you view which conditions an account satisfies for a specific health score?

In Vitally, you can expand the breakdown of a specific health score to see which conditions the account satisfies, according to Vitally. This breakdown view provides visibility into the individual criteria met by the account.

This post is general operational information, not legal advice about notice periods or renewal clauses. It does not promise that any method keeps a customer. A review model checked the facts once, and that check can miss errors, so open the sources before you act.