The Raw Events page allows you to pull reports on raw hit data.
Hits are the name for our events that are produced with every node visit by a user. As someone traverses a funnel they have a single visitor (session) ID but create a new hit ID at every node.
It is these Hits that form the raw data in our system, are what get converted when conversions happen, and ultimately tie everything together.
Because accessing raw data is demanding for analytics databases, you can only access it and certain attributes form the Raw Events section, which has some limitations -- the main one being that any query will only return a maximum of 1000 rows.
This stops you from doing a query like "show me all hits since Jan 1st 2021 for all funnels and sources", potentially trying to return 100 million rows to the UI (which your browser won't like), and with that query taking potentially a minute to compile (meanwhile impacting all other users of the database).
To pull a report, just pick the appropriate settings and event type > click load data. As a minimum, you must pick one or more specific funnels:
Understanding the data
Raw data can be a bit confusing to understand if you're not familiar with it.
Each row is an event (such as a hit) and you can then see a lot of additional data that is not available in regular reporting -- like the hit/session IDs, user-agent, timestamp, etc.
You can turn these columns on and off by clicking the "columns" button below the table:
Here for example, I can see the hits being generated by one of our uptime monitors:
And in full:
See how there is a visit to a node type "rotator" then "lander group"? This is because our link loads the traffic node, which then routes to a lander. Each one of these nodes creates its own unique hit. You can then see the traffic source and funnel names/IDs involved for each hit, the IP address and user-agent of the requester, the resolved country, and the session ID.
If you are breaking down conversion events, you will also be able to see the external ID (click IDs from the traffic source) and conversion revenue/transaction ID.
Useful ways to debug with raw data
There are many situations where this data is helpful.
Without going into great detail, here are a few:
- Checking incoming visits from a traffic source to see if they have external IDs -- without these click IDs being passed, conversion tracking to that source likely won't work
- Checking incoming visits to see if they have the correct traffic source. If using direct linking and things come up as organic, you may not be using the correct link
- Checking if clicks become change traffic source to organic - this indicates a tracking disconnect and means if they later convert, they won't be under the correct source
- Checking suspicious clicks/visits to see if they share a common user-agent or IP
- Looking through your own test data to see if every page hit, click and conversion is logged as expected
- Checking conversions to see if they have expected transaction IDs and revenue. In normal reporting you can't show each conversion as its own row, whereas here you can