Various statistics and reports are available in Blueshift via the UI, Insights reports, Campaign APIs, or Amazon S3.
In addition to the core metrics, you can create compound metrics based on core metrics or custom goals to use in reports.
Create a compound metric
You can create a compound metric from the Campaign Dashboard. You can also create a compound metric when you create an Insights report. This compound metric will then be available for use in all reports.
To add a compound metric, complete the following steps:
- Click the gear icon on the Campaign Dashboard.
- Go to the Compound Metrics tab.
- Click Edit Metrics and then + Add Metric.
- Add a Name for the metric.
- Select Format as Ratio or Percentage.
- Select a metric for the Numerator and Denominator.
- Click Create Metric.
You can now use the compound metric in your reports.
Core metrics
The table below briefly overviews all core metrics by channel.
| Push (SDK) | Push (non-SDK) | SMS | In-app | ||
| Sends | Emails sent to ESP (not delivered yet). | Messages sent to GCM/APNS. | Messages sent to SMS Service Provider (not delivered to users). | Messages sent to SMS Service Provider (not delivered to users). | In-app messages sent to customer devices. |
Note: Sends is referred to as users in the campaign activity report and other reports exported from the platform. |
|||||
| Delivered | Emails delivered to users (not bounced). | Delivered = Sent - Bounced. | Messages delivered to users (not bounced). | In-app message payload fetched and stored locally. | |
| Impressions | Total emails opened. | Displayed in the host application. | |||
| Impressions (User Only) [info] | User-only email impressions. | ||||
| Impressions (Pre-fetch Only) [info] | Impressions recorded during pre-fetch. | ||||
| Unique Impressions | Unique users who opened an email (each user counted only once). | Displayed once per unique user in the host application. | |||
| Unique Impressions (User Only) [info] | Unique impressions from user interactions only (each user counted once). | ||||
| Unique Impressions (Pre-fetch Only) [info] | Unique impressions recorded during the pre-fetch phase. | ||||
| Clicks | All clicks on tracking links. | All clicks on tracking links. | The app opened with an event within 15 minutes of sending. | Clicks on clickable elements. | |
| Clicks (User Only) | Clicks from user interactions only. | ||||
| Clicks (Bot Only) | Clicks identified as bot interactions. | ||||
| Unique Clicks | Unique users that clicked on a tracking link. | Unique users that clicked on a tracking link. | Unique users that clicked on a tracking link. | Unique users that clicked on a tracking link (if enabled). | |
| Unique Clicks (User Only) | Unique clicks from user interactions only. | ||||
| Unique Clicks (Bot Only) | Unique clicks identified as bot interactions. | ||||
| Dismiss | The customer closed or dismissed the in-app message. | ||||
| Visits | New session from a campaign click. | New session from a campaign click. | New session from a campaign click. | ||
| Orders | Orders via a campaign. | Orders via a campaign. | Orders via a campaign. | Orders via a campaign. | |
| Activations | First purchase via a campaign. | First purchase via a campaign. | First purchase via a campaign. | First purchase via a campaign. | |
| Revenue | Total revenue from the campaign. | Total revenue from the campaign. | Total revenue. | ||
| Bounces | Hard bounces (permanent rejection). | Invalid tokens (removed from future sends). | Invalid tokens (removed from future sends). | Bounced messages. | |
| Soft Bounces | Temporary delivery issues (retries up to 72 hrs). | ||||
| Spam Reports | Reported as spam by a user. | ||||
| Unsubscribes | Users who unsubscribed via link. | Unsubscribes via specific keywords (e.g., STOP). | |||
| Device Sends | Devices a push message was sent to. | Devices a push message was sent to. | Devices an in-app message was sent to. | ||
| Device Bounces | Devices that could not receive the push message. | Devices that could not receive the push message. | Devices that could not receive an in-app message. | ||
| Drops | Messages dropped by ESP for sender protection. | ||||
Understanding pre-fetch impressions
Pre-fetch impressions are recorded when email clients or systems automatically preload images, such as tracking pixels, without any action by the recipient. This is commonly done for caching purposes and does not represent genuine user engagement.
In contrast, user-only impressions are recorded when recipients actively engage with an email, such as opening it in their email client or viewing it in a browser.
| Type of impression | Description | Example |
|---|---|---|
| Pre-fetch impression | Recorded when the user-agent string contains only basic information (e.g., Mozilla/5.0) without device, OS, or browser details, indicating automatic image preloading. | An email client like Apple Mail preloads images automatically when receiving an email, but the user doesn’t open it. |
| User-only impression | Recorded when the user-agent string includes detailed device and browser information, indicating actual user engagement. | The recipient opens an email in their client or clicks to view it in a browser, actively loading the images. |
Note: Certain email privacy features or proxy services (e.g., Apple’s Mail Privacy Protection) can obscure the user-agent string, making it harder to distinguish pre-fetch impressions from user-only impressions.
To explore the technical aspects in greater detail, visit our blog post: iOS 15 Updates: Keep Adapting to Mail Privacy & Proxies (under the section - Some Technical Background Info).
Key insights on reporting:
- Blueshift uses the HyperLogLog algorithm to estimate unique impressions and clicks. This algorithm is accurate at 98%, which means that results will be within a 2% error margin.
- For a campaign, if you consider the uniqueness counts for a single day, these are calculated independently for each day, without any influence from other days. For example, for a campaign, one user clicks on the same message in their inbox every day for 7 days. If you generate a daily time series, you will see this unique click in the report for each day. But you will see a single click if you generate a report for 7 days.
Comments
0 comments