> For the complete documentation index, see [llms.txt](https://doc.batch.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://doc.batch.com/guides-and-best-practices/batch-ai/how-to-retain-your-best-customers-before-its-too-late.md). # How to Retain Your Best Customers Before It's Too Late ## What Is the At-Risk Top Customer Score? Batch AI Predict's **At-risk top customer** score identifies, among your historically high-value customers, those whose engagement signals are decreasing. The objective is to act before a potential decrease in their spends. The output is a float between 0 and 1: 1 is for the highest risk and 0 for the lowest. The score is **recalculated on a regular basis**. New customers cross the risk threshold at each recalculation, so retention actions can be triggered on a continuous basis rather than in periodic batches. {% hint style="warning" %} This score is designed specifically for **top-value customers**. It is computed across your entire database, but is most meaningful when applied to customers who have historically been among your highest spenders or most engaged profiles. It is not a generic churn score for your full base. {% endhint %} ## Why Use It? Most loss of momentum is invisible until it is too late. By the time a top customer has stopped purchasing or engaging, re-engagement becomes significantly harder and more expensive. With the At-risk top customer score, you can: * **Detect early degradation signals** among your most valuable customers, weeks before they go fully inactive. * **Trigger proactive retention actions** (personalized offers, VIP communications, or relationship outreach) while the customer is still reachable. * **Avoid wasting retention budget** on customers who are not at risk, or on customers who are too far gone to recover. ## How to Use This Score in Batch ### Trigger Mode: Automate Retention When a Customer Crosses the Threshold The recommended setup. Because the score updates regularly, new at-risk customers enter the segment continuously. A trigger-based automation ensures they receive a retention message as soon as they cross the threshold, without manual intervention. **How to set it up:** 1. Go to **Automations** in your Batch dashboard and click on **New Automation**. 2. Create a new automation with an **Attribute Change** trigger: * Trigger condition: `Predictive score [At-risk top customer] crosses above 0.5` (or your configured threshold) 3. Define your retention message: a personalized email, a push notification, or a coordinated sequence. 4. Add a **capping rule** to ensure customers are not re-entered into this automation too frequently (for example, no more than once every 30 days). 5. Activate the automation. From that point on, every customer who becomes at risk automatically receives the retention message. **Recommended test**: Use a **Control Group** on your automation; exclude a random portion of at-risk customers from the retention flow and compare their behavior over 30-90 days against the customers who received the message. This gives you a clean measure of the retention uplift generated by the automation. ### Campaign Mode: Manual Periodic Campaigns If you are not yet ready to implement a trigger-based automation, you can run regular manual campaigns targeting at-risk top customers. **How to set it up:** 1. Create your campaign in Batch. 2. In the targeting step, apply a condition: `Predictive score [At-risk top customer] ≥ 0.5` (or your configured threshold). 3. Optionally combine with a filter to restrict to historically high-value customers (for example, customers in the top 20% by historical revenue). {% hint style="info" %} If you plan to run this campaign regularly, save the audience as a named segment (for example, `[AI] Top clients at risk`) so you can reuse and monitor it easily over time. {% endhint %} {% hint style="warning" %} In campaign mode, customers who become at risk between two sends will not be reached until the next scheduled campaign. The trigger mode is more effective for this reason. {% endhint %} ## Real-World Use Cases | **Industry** | **Example use case** | | ------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Retail** | VIP customers whose purchase frequency is declining: send a proactive "reserved for our best customers" push or email before they stop buying altogether | | **Media & Entertainment** | Premium subscribers whose weekly engagement is dropping: send a curated content digest or exclusive early-access communication to re-engage them | | **Travel** | Gold-tier travelers who have not booked in over a year: reach out with priority access to a private sale or a personalized destination suggestion | | **Retail (subscription)** | Loyalty program members showing early signs of disengagement: trigger a renewal incentive or a loyalty milestone message before they lapse | *** ## Related Documentation * [How to Measure the Performance of Batch AI Predict Scores](https://doc.batch.com/guides-and-best-practices/batch-ai/how-to-measure-the-performance-of-batch-ai-predict-scores) * [Segments](https://doc.batch.com/getting-started/features/customer-engagement-platform/profiles/segments) * [Orchestration Targeting](https://doc.batch.com/getting-started/features/customer-engagement-platform/orchestration/targeting) * [Orchestration Analytics](https://doc.batch.com/getting-started/features/customer-engagement-platform/analytics/orchestration-analytics) * [Conversion Goal dedicated insights](https://doc.batch.com/getting-started/features/customer-engagement-platform/analytics/orchestration-analytics#conversion-goal-dedicated-insights) * [Omnichannel Trigger Automations](https://doc.batch.com/getting-started/features/customer-engagement-platform/orchestration/trigger-automations?q=target#analysing-automation-with-analytics) --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://doc.batch.com/guides-and-best-practices/batch-ai/how-to-retain-your-best-customers-before-its-too-late.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.