What are Weekly Insights by Cleeng AI Assistant
Weekly Insights is one of the Cleeng AI Assistant's agents. It acts as your dedicated subscription analyst, delivering data-driven insights directly within the Cleeng dashboard.
Every week, it reviews your metrics from the past two weeks to identify specific patterns, risks, and opportunities that might otherwise get lost in the noise of raw data.
Instead of just showing you static numbers, it translates technical data (like payment gateways, offer codes, and offerIDs) into clear, readable descriptions - converting complex codes into plain terms like Early-stage or Returning Subscribers.
Note: This feature is currently in Early Access. This means it's ready for real-world use while we gather user feedback to enhance its functionality and address any minor issues.
How to see my Weekly Insights
You can access Weekly Insights directly through the Cleeng AI Assistant chat in your dashboard. Type "Show my Weekly Insights" to view your latest report.
How to interpret my Weekly Insights
The Weekly Insights report is structured to guide you from high-level observation to specific action. Here is how to read the four sections of the report:
- Key findings: This is your at-a-glance summary. It highlights the most significant behaviors occurring across your platforms, such as revenue concentration in a specific app or a sudden spike in churn for a specific payment method.
- Positive trends: Look here to see what is working. The AI highlights segments showing growth, stability, or improved retention, helping you understand where your revenue is most secure.
- Negative trends: This section acts as an early warning system. It flags underperforming segments - such as high churn rates on specific plans or failed payments, so you can intervene quickly.
- Key segments to target: This is the most actionable part of the report. The AI selects the top 3 specific customer groups you should focus on immediately, providing a What, Why, and Recommended Action for each.
How to read segment names
Each segment name in your Weekly Insights describes a specific group of subscribers. The AI builds these names by combining up to six subscriber characteristics, so a single segment name tells you several things about that group at once.
Naming pattern
Segment names follow this general pattern:
[Payment method] [Distribution channel] [Subscription tenure] [Subscriber type] ([Offer ID or Country])
A real example: Card Web Early-stage First-time Subscribers (S12345678)
Not every segment includes all dimensions. The AI only includes the characteristics that are most relevant for that particular finding. For instance, a segment might appear as just Roku platform Long-term First-time Subscribers if offer and country are not the distinguishing factors.
Segment dimensions explained
1. Payment method
The billing method used by subscribers in this group - for example, Card, iOS, Android, PayPal, or Roku. The AI automatically selects the most popular payment methods for your account (typically the top 4), so you will only see methods that are common enough in your data to form a meaningful segment.
2. Distribution channel (platform)
The platform or storefront where the subscription was purchased - for example, Web, App Store, or Roku platform. As with payment methods, only the most popular channels for your account will appear.
3. Subscription tenure
How long subscribers in this group have been continuously subscribed. This is a binary split based on your account's data:
| Label | What it means |
| Early-stage | Subscribers whose subscription duration is below your account's average. These are your newer customers - they carry the highest churn risk, and retention efforts here have the greatest impact on long-term value. |
| Long-term | Subscribers whose subscription duration is above your account's average. These are your more established, loyal customers who have survived the early churn window. |
4. Subscriber type
Whether this is the subscriber's first time on your platform, or if they have subscribed before:
| Label | What it means |
| First-time Subscribers | Customers who have only ever had one subscription on your platform. This is their first (and current) subscription. |
| Returning Subscribers | Customers who have had more than one subscription historically - meaning they previously churned and came back. Useful for evaluating whether your win-back efforts are working. |
5. Offer ID (when shown)
Some segments include an offer identifier in parentheses, e.g. (S12345678). This refers to a specific subscription offer (plan) on your account. When an offer ID appears, it means the insight is specific to subscribers on that particular plan - not your entire base.
When the offer ID is not shown, the insight applies across all offers for that segment combination.
6. Country (when shown)
Some segments include a country code in parentheses, e.g. (GB). This means the finding is specific to subscribers from that country. Country appears when there is a geographically concentrated trend that is meaningful enough to surface separately.
Examples
Here is how to read three real segment names from an actual Weekly Insights report:
Example 1: "Card Web Early-stage First-time Subscribers (S12345678)"
| Dimension | Value | What it tells you |
| Payment | Card | Paying by credit/debit card |
| Channel | Web | Purchased via the website |
| Tenure | Early-stage | Below-average subscription duration (newer customers) |
| Type | First-time subscribers | First subscription on the platform |
| Offer | S12345678 | Specific to this offer/plan |
How to read it:
These are new web signups on your primary card-billing offer. If this segment appears under ⚠️ Negative Trends with declining new subscriptions and rising churn, it signals a problem with your main acquisition funnel - consider improving onboarding, addressing payment failures, or testing early-life retention offers.
Example 2: "iOS App Store Long-term First-time Subscribers"
| Dimension | Value | What it tells you |
| Payment | iOS | Paying through Apple |
| Channel | App Store | Purchased via the App Store |
| Tenure | Long-term | Above-average subscription duration (loyal customers) |
| Type | First-time Subscribers | First subscription on the platform |
How to read it:
These are loyal Apple subscribers who have been with you longer than average. If Weekly Insights reports 0% churn and stable revenue here, this is your foundation - protect it. If anything starts to shift, investigate immediately because this cohort is high-value.
Example 3: "iOS App Store Long-term Returning Subscribers (GB)"
| Dimension | Value | What it tells you |
| Payment | iOS | Paying through Apple |
| Channel | App Store | Purchased via the App Store |
| Tenure | Long-term | Above-average subscription duration |
| Type | Returning Subscribers | Previously churned and resubscribed |
| Country | GB | United Kingdom |
How to read it:
These are UK-based Apple subscribers who previously left but came back and have now stayed longer than average. A positive trend here (e.g. +13% revenue growth, 0% churn) is strong evidence that your win-back strategy is working specifically for UK iOS users.
Clicking through to segment details
Each segment mentioned in your Weekly Insights includes a clickable link. Clicking it opens the Segment Builder, pre-filtered to that exact subscriber group. From there you can:
- Explore the underlying data in detail
- Export the subscriber list
- Launch a targeted retention or engagement campaign
Important note on data scope & logic
-
It is NOT a summary of your entire base
Weekly Insights does not provide aggregate totals for your whole business (e.g., Total Active Subscribers or Total Monthly Revenue). Instead, the AI model automatically filters your data to find only the specific segments that are statistically significant. -
How segments are prioritized
If you see a segment listed, it is because the AI has determined it is one of the most important groups for you to look at right now based on four specific criteria:- Business impact: Segments with significant revenue size, growth potential, or high risk.
- Actionability: Areas where specific interventions (like a campaign or fix) can drive results.
- Strategic value: Segments representing long-term value or market expansion opportunities.
- Urgency: High churn risks or rapidly declining trends that need immediate attention.
-
Limitations & best use cases
To ensure you are reading the data correctly, please keep the following in mind:- Subscribers vs. passes: The model is optimized to analyze recurring revenue and retention patterns. It works best for subscription plans (monthly/annual) and is less effective for one-time purchases like Passes or PPV.
- Data volume: The reliability of the findings depends on having a steady flow of data. In weeks with low transaction volumes, findings may be less precise or subject to statistical noise, as small changes in user behavior can disproportionately skew the percentages.
- Segment granularity: You may notice that some segments include more dimensions (e.g., "Card Web Early-stage First-time Subscribers (S12345678)") while others include fewer (e.g., "Roku platform Long-term First-time Subscribers"). This is intentional - the AI Analyst surfaces the level of detail that is most useful for each finding. A broader segment appears when the trend applies across multiple sub-groups.