Propensity modeling ranks your existing donors by how likely they are to take a specific action, such as starting recurring giving or responding to a campaign theme. Use it to target the right donors with the right ask at the right time.
Use this article when
- You want to understand what propensity modeling is and how it works in Avid.
- You need to decide whether a propensity model or a lookalike audience fits your goal.
- You are ready to create or refresh a propensity model from Pathways > Manage Audiences.
Applies to and prerequisites
| Requirement | Details |
|---|---|
| Permission | Pathways > Edit |
| Segment type | Owned segment built on a constituent data blend |
| Minimum seed size | 1,000 constituents |
If you do not see Create a Propensity Model in the action menu, confirm that your role includes Pathways > Edit permission and that the seed segment meets the minimum size requirement.
What is propensity modeling?
Propensity modeling predicts which donors are most likely to take a specific action in the near future.
Avid examines what past donors looked like before they took an action (for example, starting recurring giving), then finds donors in your file who currently look similar. Each donor receives a probability score so you can prioritize outreach.
Technical details
A propensity model is a supervised classification model. It trains on labeled outcomes (donors who did vs. did not take the action) and produces a probability score for each donor.
Avid uses point-in-time feature engineering to avoid data leakage. Training features come from behavior before the qualifying action. Scoring features come from the donor's current profile.
Propensity vs. lookalike: when to use each
| Lookalike model | Propensity model | |
|---|---|---|
| Primary goal | Find new people | Prioritize existing donors |
| Best for | Acquisition | Retention, upgrades, cultivation |
| Population | Prospects and non-donors | Current and lapsed donors in your CRM |
| Output | New audience expansion | Ranked list of who to ask now |
| Avid feature | Create a Lookalike Acquisition Audience | Create a Propensity Model |
Use lookalike when you need new names. Use propensity when you need better decisions on your current file.
When to use propensity modeling
1. Recurring donor conversion (best first model)
Question: Who is most likely to start recurring in the next 60-90 days?
Recurring conversion is strongly behavior-driven. Leading indicators like recent cadence, frequency, and consistency make this a reliable model. Teams use it to target sustainer upgrade campaigns, trigger recurring email journeys, and prioritize donors for monthly ask calls.
2. Appeal theme affinity
Question: Who is most likely to respond to a specific theme?
Donors often have repeatable emotional and category preferences (disaster relief, education, healthcare, faith-based, animal welfare). Appeal behavior tends to stay consistent across campaigns. Teams use it to build theme-specific audiences, personalize messaging by affinity, and expand campaign reach with better relevance.
3. Next best ask for active donors
Question: What is this donor most likely to do next?
Possible next actions include starting recurring, responding to a specific theme, or upgrading giving. This model uses real transaction and channel behavior to improve ask selection and message fit.
4. Lapsed donor re-engagement
Question: Which lapsed donors are likely to return soon?
Recency and historical giving pace are strong signals. This model helps focus limited reactivation budget on recoverable donors.
5. Channel optimization
Question: Which donors respond best to which channel?
Channel preference is usually stable over time. This model improves response rates while reducing waste.
When not to use propensity
Use a different approach when:
- No clear target action exists -- for example, "good donor" with no measurable event.
- Outcome volume is too small -- not enough historical positives to train on.
- You need brand-new prospects -- use the Create a Lookalike Acquisition Audience option instead.
- You need very long-horizon predictions only -- vague multi-year outcomes with no near-term behavior signal.
- Data is stale or sparse -- recent behavior is missing, which makes scores unreliable.
Where to start
If you build only two models first, start here:
- Recurring propensity (90-day horizon) -- highest ROI and easiest to explain.
- Appeal theme affinity -- directly improves targeting and messaging.
Together these unlock better targeting, message match, and timing.
Example: combining two models for tighter targeting
Create a propensity model
- Go to Pathways > Manage Audiences.
Find or create the seed audience that defines the target action.
Example seed: "donors who started recurring in the last 24 months."
- Click the action menu (three-dot icon) on the segment row.
Select Create a Propensity Model.
Enter your Target Audience Size.
This is the number of accounts to include in the resulting audience.
- Click Create and wait for processing.
- When complete, Avid creates a new audience named Propensity Model of {Seed Segment Name}.
- Open that audience and use it in Playbooks and Pathways pushes.
Refresh an existing model
Refresh when you want the same modeling logic with the latest donor behavior.
- Go to Pathways > Manage Audiences.
- Find the seed audience that already has a propensity model.
- Click the action menu (three-dot icon) on the segment row.
- Select Create a Propensity Model again.
- Enter the new target audience size (or keep the same).
- Click Create.
- Wait for processing to complete, then review the updated audience membership before launch.
Refresh on a regular cadence (monthly or campaign-cycle based) so scores reflect current behavior.
Other modeling options in Manage Audiences
Avid offers several audience modeling tools from the same action menu. Each serves a different purpose:
| Option | What it does | Best for |
|---|---|---|
| Create a Propensity Model | Scores existing donors by likelihood to take a specific action | Retention, upgrades, re-engagement |
| Create a Lookalike Acquisition Audience | Finds new prospects who resemble your best donors | Acquisition campaigns |
| Create an Analysis Audience | Builds a behavioral analysis audience with appeal code and date filters | Detailed donor behavior analysis |
| Create a Persona Audience | Adds a value proposition description to guide messaging | Copy and creative personalization |
Use propensity audiences with Pathways
From Manage Audiences, you can:
- Sync to connected destinations for campaign execution.
- Combine propensity audiences with additional segment filters.
- Route to specific channels based on response patterns.
- Refresh models as behavior changes.
- Measure lift against baseline audiences in Insights.