Account Scoring
Scoring
What is Account Scoring?
Account scoring in HockeyStack predicts how likely each account is to reach a specific goal — such as closing a deal, booking a demo, or becoming a qualified lead. Instead of relying on gut feel or static rules, HockeyStack uses a machine learning model that learns directly from your historical data to surface the accounts that matter most.
The model studies patterns in accounts that have already reached that goal and applies those patterns to every account in your pipeline. The result is a score from 0 to 100 for each account, updated automatically, so your team always knows where to focus.
How the Score Works
Learning from your data
HockeyStack analyzes the behavior and characteristics of accounts that have already hit your chosen conversion goal. The model identifies which engagement patterns — like visiting certain pages, attending events, or downloading content — are most predictive of conversion. It then scores every account based on how closely its activity matches those winning patterns.
Because the model is trained on your data, the scores reflect what actually drives conversion in your business — not generic industry benchmarks.
What the model looks at
Depending on the type of score you create, the model considers:
Behavior signals: Website visits, content engagement, email interactions, event attendance, ad clicks, and other touchpoints your team tracks.
Score range
Every account receives a numerical score between 0 and 100. A higher score means the account more closely matches the engagement patterns of accounts that have already converted. A score of 0 means there isn't enough data to score the account yet.
There are no fixed categories — the score is a continuous number, so you can set your own thresholds based on what works for your team (e.g., "prioritize anything above 65" or "nurture accounts between 30 and 60").
Scores stay current
Scores are recalculated regularly so they reflect the most recent activity. If an account goes quiet, a recency decay gradually lowers its score over time — ensuring that stale engagement doesn't keep an account artificially high. Conversely, a burst of new activity will push the score up.
0–30 days
Full score (no decay)
31–60 days
Slight reduction
61–90 days
Moderate reduction
91–120 days
Significant reduction
120+ days
Maximum reduction
Scoring factors
Rather than a black-box number, each score comes with scoring factors — a breakdown of the specific behaviors and attributes that contributed to the score. For example, you might see factors like:
"Visited pricing page 4 times in the last 14 days"
"Attended webinar"
"Multiple stakeholders engaged"
These factors make it easy to understand why an account scored the way it did and give your team more concrete data for next steps.
Setting Up a Score
You can create and manage scores from the Atlas > Scoring section (Atlas accessible from the sidebar). You can set up multiple scores — for example, one for "Closed Won" deals and another for "Demo Booked" — each tailored to a different conversion goal.
Step-by-step setup
Creating a new score takes just a few minutes. A guided wizard walks you through each step:
1. Name your score
Give it a descriptive name your team will recognize, like "Deal Won Score" or "Demo Readiness Score."
2. Choose the score type
Pick the kind of signals the model should focus on:
Behavior Score — Tracks engagement signals like page visits, content downloads, email opens, and event attendance. Best for measuring intent through actions.
Fit Score — Evaluates how well a prospect matches your ideal customer profile based on firmographic data from your CRM. Best for qualifying leads based on company attributes. (Coming soon)
3. Choose the entity level
Decide whether to score at the company (account) level or the person (contact) level:
Company Score — Aggregates all activity and attributes from the entire organization into a single score.
Person Score — Scores individual contacts based on their personal engagement and attributes. (Coming soon)
4. Set a conversion goal
Select the goal that defines "success" for this score. This is the outcome the model will learn to predict. Choose from the goals you've already defined in HockeyStack — for example, "Deal Won," "Demo Booked," or "MQL Achieved."
The model will study accounts that have already reached this goal and learn what they have in common.
5. Select your account list(s)
Choose which accounts the model should score. You can select one or more Sales Views that define the set of accounts you want to include. This lets you scope the score to the right set of accounts — for example, only accounts in a particular segment, region, or pipeline stage.
6. Scoring details
For Behavior Scores:
Touchpoint Properties — Select which engagement properties (like UTM source, campaign, or channel) to include as scoring signals. Up to 3 properties.
Score Breakdowns — Choose dimensions to segment the model by (like industry or company size). This trains separate models per segment for more precise scoring. Up to 3 breakdowns.
7. Field name for score sync to CRM
Provide custom field/property names to use when syncing this score to CRM (Salesforce or Hubspot). Ensure that the field visibility settings are set up correctly by the CRM admins.
8. Use in Blueprints toggle
When enabled, this score will be used to power the Marketing Blueprints feature — surfacing conversion patterns, funnel insights, and next steps based on this score's data.
After you save
Once you save, the model begins training on your data. You'll see a "Running" status on the score while the pipeline is working. When it's done, the status changes to "Completed" and scores are available across the platform. This typically takes a few minutes depending on data volume.
Managing your scores
From the Scoring Configuration screen, you can:
Edit any existing score — update the name, conversion goal, audience, or configuration.
Delete a score you no longer need.
Create additional scores to track different conversion goals or audiences side by side.
Enable for Blueprints for Marketing Blueprints feature.
Each score in the list shows its name, type (Behavior or Fit), entity level (Company or Person), conversion goal, and current processing status.
Companies List Page
When you navigate to the Companies page in the sidebar, you'll see Score(s) column alongside each account. This score reflects the account's current engagement level relative to your scoring configuration.
What you'll see
Numeric score (0–100): The account's current score.
Multiple scores: All scores configured for that sales view.
Filter by score
Use the built-in filters to narrow your list based on score values. You can filter by score range to focus on the accounts that matter most — for example, showing only accounts above a certain threshold.
Use the built-in filters to narrow your list by score range — for example, showing only accounts scoring above 60 to focus on your strongest opportunities, or filtering to 20–50 to find accounts worth nurturing.
Show or hide the scores columns
The score column can be toggled on or off from the column visibility settings on the companies list. If your team doesn't need to see scores in this view, you can hide the column to keep the table clean — and bring it back anytime.
Sort
Click the Score column header to sort accounts from highest to lowest (or vice versa), so your best opportunities always appear at the top.
Single Company Page
When you click into an individual company, the Overview tab includes a dedicated summary that gives you a deep look into that account's scoring.
Score display
At the top of the Intent card, you'll see:
The score (0–100) displayed prominently. Higher scores indicate stronger alignment with conversion patterns.
A color-coded icon provides a quick visual cue — the warmer the color, the higher the score.
Score trend chart
Below the score, a line chart shows how the account's score has changed over the past few months. This makes it easy to see:
Whether engagement is trending up or down
When spikes or drops occurred (and correlate them with campaigns or events)
Whether the account is gaining or losing momentum
Scoring factors
Below the chart, you'll find the Scoring Factors section — a list of the specific signals that contributed most to this account's score, ranked by impact. Each factor shows:
What happened (e.g., "Pricing page visits," "Webinar attendance," "Multiple stakeholder engagement")
How much it contributed to the score (shown as a weight)
This breakdown helps reps understand why this account scored the way it did, so they can tailor their outreach. For example, if "Pricing page visits" is the top factor, a rep might lead with pricing-related content.
Last scored
A timestamp shows when the account was last scored, along with the total number of tracked actions. You can also manually trigger a rescore by clicking the rescore button if you want an immediate update after new activity.
Syncing Scores to Your CRM
The real power of scoring comes from acting on it — and the fastest way to do that is by syncing scores directly to your CRM so they're available alongside your deals, tasks, and workflows.
How to enable CRM sync
Score syncing is configured per Sales View. To set it up:
Open the Sales View you want to sync from.
Click the configuration dropdown (gear icon).
Find the Sync to CRM section.
Toggle Active to enable syncing.
Add one or more destinations — Salesforce or HubSpot.
Supported CRM platforms
Salesforce
Fully supported (including Salesforce Sandbox)
HubSpot
Fully supported (including HubSpot Sandbox). Select the specific HubSpot account when you have multiple connected.
You can sync to multiple destinations at once — for example, both Salesforce and HubSpot — if your team uses more than one CRM.
What gets synced
When CRM sync is enabled, HockeyStack writes the following fields to your account records:
Intent Score
The numeric score (0–100)
Intent Score Status
The trend — New, Surging, Heating, Cooling, or Neutral
Last Scored At
Timestamp of the most recent scoring run
Score Reasoning
A summary of the top factors that contributed to the score
These fields are created automatically on your CRM account/company objects the first time they sync.
Using scores in CRM workflows
Once scores are in your CRM, you can build workflows and automations around them. Common use cases include:
Alert reps when an account crosses a score threshold (e.g., score goes above 60).
Auto-assign accounts to reps based on score range.
Trigger sequences or outreach cadences when a score rises past a threshold.
Pause outreach when an account's score drops below a threshold.
Build reports and dashboards in your CRM that segment accounts by score range.
Sync frequency
Scores sync automatically on a regular cadence aligned with your scoring frequency. You can also configure the scoring frequency (how often accounts are re-scored) from the same configuration dropdown in your Sales View.
Setting the frequency to Off will disable both rescoring and CRM sync for that view.
Accounts That Don't Receive a Score
Some accounts are excluded from scoring:
Closed-Won deals: Once a deal is marked as won, HockeyStack stops scoring that account — since it's already converted, a predictive score is no longer relevant.
No verified domain: Accounts without a verified company domain cannot be scored, since engagement data is tied to the domain.
Best Practices
Set up properties accurately. Goals, Touchpoint Properties, Breakdowns are crucial to the scoring models. Ensure that they are correctly set up.
Provide sufficient training data. Add the sales view that has the maximum number of companies. Usually this would be "All companies with domain".
Provide the right Touchpoint Properties. We have found properties like "Channel" and "Asset Type" to be the most useful.
Start with your primary conversion goal. Create your first score around the outcome that matters most to your team (e.g., "Closed Won") and validate it against recent deals before expanding.
Check against known outcomes. Look at recently closed deals and see if the model would have scored them highly. This builds confidence in the score before you roll it out broadly.
Use breakdowns for precision. If your business serves different segments (e.g., Enterprise vs. SMB), add a breakdown dimension so the model trains separate patterns for each segment.
Combine score and trend. A mid-range account that's Surging may be more urgent than a high-scoring account that's Cooling. Use both signals to prioritize.
Keep your audience scoped. Use Sales Views to score only the accounts that matter for a given team or motion — this keeps scores relevant and avoids noise.
Review factors regularly. Scoring factors tell you what's working. If "Pricing page visits" consistently appears as a top factor for converted accounts, that's a signal to drive more traffic there.
Involve your team. Share scoring factors and methodology with your sales and marketing teams so everyone understands what the numbers mean and trusts the score.
Last updated