What's new
The GraphComment back-office analytics dashboard has been completely revamped to provide precise analysis periods and reliable data. Here are the key improvements:
Calendar-based periods
The analysis windows now use calendar-based periods (ISO weeks or full months) instead of sliding windows ("last 7 days", "last 30 days") via a new date picker. A 12-month trend table is available to track monthly performance.
Accurate unique visitor counts
The unique visitor count is now exact for calendar-based periods. Double-counting the same visitor across days has been eliminated, preventing inflated figures (previously, figures could be overstated by 1.5× to 5×).
Real engagement measurement
A new tool measures engagement server-side using five reliable and auditable sensors:
- Readers who actually read comments.
- Engaged readers (reading time, interactions).
- Average reading time.
No browser-side declarative measurements are used.
How to use it
- Access the dashboard: Log in to your GraphComment back-office and go to the "Statistics" section.
- Select a period: Use the new date picker to choose an ISO week or calendar month as your analysis period.
- Analyze the data: Review key metrics such as page views, engaged readers, and comments. Unique data is displayed without double-counting.
- Filter results: Apply filters by article, section, or author to refine your analysis.
Use cases
- Monthly tracking: Compare performance between two consecutive months using fixed periods.
- Weekly analysis: Identify weekly trends to adjust your publishing strategy.
- Reporting: Export reliable data for partners or advertisers.
FAQ
Is this feature available on all plans?
Yes, the revamped audience statistics are available for all GraphComment customers, including the free plan.
Do I need to activate anything to access it?
No, the new features are enabled by default. Simply log in to your back-office and navigate to the "Statistics" section.
Can I export the data?
Yes, data can be exported in CSV format from the dashboard. You can filter periods and metrics before exporting.