Product updates for May-June’23

1. Try automated marketing reporting: self-service free trial available now

Now, you can try SegmentStream’s Automated cross-channel marketing reporting without talking to sales.

Here’s how to get started:

• Log in to start a 14-day free trial
• Integrate SegmentStream using Google Tag Manager
• Connect ad platforms
• Set up conversions and reports
• Enjoy 14 days of seamless CPA/ROAS reporting in one interface

2. Model reliability evaluation

With this feature, we automatically evaluate the eligibility of each simple conversion action for modelling.

If a conversion is unlikely to yield accurate results, modelling will be disabled in the interface. This prevents inaccurate predictions and ensures more reliable insights.

3. New project setup flow

1. Connecting Google Analytics has become easier:

Users can choose their preferred connection method - either manually (gtag.js) or using Google Tag Manager (GTM) - and follow straightforward instructions for a seamless setup.

Additionally, users can now monitor the real-time setup status through the user interface, ensuring a smooth and hassle-free integration process.

2. SegmentStream-managed data warehouse:

For customers without their own data warehouse, we provide the option of a SegmentStream-managed DWH. You don’t need to set up your own infrastructure and use ready-made integration instead.

3. Direct BigQuery integration:

Easily connect existing BigQuery projects and create new datasets right from the SegmentStream interface, eliminating the need for technical setup through Google Cloud and providing a user-friendly experience for all customers.

4. New in the Notification Centre

1. Anomaly detection

Receive notifications about detected anomalies in conversions to maintain the health of ML models.

2. Value parameter alerts

Get notified when conversions do not pass the value parameter for accurate tracking.

5. Modelling of combined conversions

With the latest update, SegmentStream allows to use different conversion actions for training the ML model.

This is helpful for companies with multiple desired actions of varying value. For example, an e-commerce store can evaluate traffic sources by considering all valuable events like purchases, calls, and price requests together, rather than separately.