Skip to main content

Analyze stream traffic in 15-minute intervals

C
Written by Customer Success

The all-streams traffic export has an overview file for each stream. It splits viewing into 15-minute intervals so you can see when each stream had the most viewers and the most watch time.

Before you start

  • You need to be an organization admin or an event admin.

  • Streams must be on. If Content > Streams shows "You haven't enabled streams yet!" or "You have disabled streams!", select Enable streams or Re-enable streams. The Streams, Analytics and Settings tabs appear.

  • Brella has no Google Sheets export. The files are CSV files that any spreadsheet app can open, including Google Sheets and Excel.

Where to find it

Content > Streams > Settings > Export stream traffic data.

Export the overview file

  1. Open Content > Streams, then the Settings tab.

  2. Next to Export stream traffic data, select Export .zip.

  3. The message "Your export will be delivered by email shortly." confirms the request.

  4. Open the zip from the email. It has 3 files for each stream of the event, hidden streams included, numbered in the order the streams were created. The overview file is in the folder whose name ends in -stream-overview, and its name ends in -overview.csv.

The other 2 files per stream are in the folder whose name ends in -stream-visitors: one ending in -interactions.csv and one ending in -visitors.csv. See Streams traffic.

Export (.csv) on the Analytics tab exports one stream without the overview file. The overview file comes only with Export .zip.

Streams Settings tab showing Use streams and Export stream traffic data with the Export .zip button

Columns in the overview file

Each row is one 15-minute interval.

Column

What it means

date

Date of the interval, in the event's time zone.

interval_start

Start time of the interval, for example 09:00.

interval_end

End time of the interval, for example 09:15.

unique_viewers

Attendees who watched during the interval. Each attendee counts once per interval.

companies

Different company names among those viewers. Never higher than unique_viewers.

total_watch_time

Seconds watched by all viewers in the interval. Only the part of each viewing that falls inside the interval counts.

avg_viewers

total_watch_time divided by 900 (the seconds in 15 minutes), rounded to 2 decimals. It is the average number of people watching at any moment.

total_view_count

How many times viewers opened the stream during the interval.

Example for avg_viewers: one person watches the first half of an interval and another watches the second half. total_watch_time is 900 and avg_viewers is 1.

Company names count as different when they are spelled differently. A blank company name also counts.

Which intervals are included

  • Intervals run from 00:00 on the event's start date to 24:00 on its end date, in the event's time zone.

  • Only intervals with at least one viewer have a row.

  • Viewing outside the event dates is not in the overview file. It is in the interactions and visitors files.

  • A viewing without a recorded end uses an estimated duration. Export after viewing has finished when comparing final watch time.

  • If nobody has watched a stream, its overview file is empty.

Find the busiest moments

  1. Import the overview CSV into your spreadsheet app. Choose comma as the separator if asked.

  2. Check that the column names occupy separate cells in the first row.

  3. Select the whole table and mark the first row as headers before sorting. This keeps each interval’s values together.

  4. Sort unique_viewers from largest to smallest to find the busiest intervals.

  5. Sort total_watch_time for total viewing time, or avg_viewers for the average number watching at once.

  6. Sort by companies to see when the most companies were watching.

  7. Sort by total_view_count to see when viewers opened the stream most often.

  8. Compare the interval times with your schedule in the event’s time zone.

For example, suppose the sorted file shows these intervals on the same date:

interval_start,interval_end,unique_viewers
10:00,10:15,80
10:15,10:30,55

The first interval reached more people. If the keynote ran from 10:00 to 10:30, both intervals belong to it. Do not add 80 and 55 to count keynote viewers: the same person can appear in both intervals.

If two sessions overlap an interval, the row cannot tell you which session each person watched. Compare streams over the same interval and review session analytics separately.

Limits

  • The interval length is fixed at 15 minutes.

  • Only signed-in attendees who watch in the mobile app or the web app are counted. Viewers on the stream provider's own page are not.

  • The overview file always covers the whole event. It has no time frame setting.

Related articles

Did this answer your question?