As I posted on part 1 of the series, what really attracted me to Tealeaf was the data. A lot people see tealeaf purely as a session replay tool, and that can’t be farther from the truth. In order to create the replay, Tealeaf stores the user’s request/response data (all data passed to and from the web servers), as well as ALL interactions on the page (like clicks, data entry, etc. — often referred to as the UI data). A part of our clients also track API calls (both internal/external) and some push in data from other sources (like customer support, survey data, etc). So, this is all interesting for customer replay, but extremely interesting for Analytics.
So, with all this data stored inside Tealeaf, one of the more powerful features of Tealeaf is the session search. The power here is the ability to search through Months of data looking for something specific that may have happened on the web site. If a new error appears and is only discovered a week later, we can search for the error message and quickly see all the sessions that experienced the error. We may not have “evented” the error (tealeaf can look for specific scenarios), but all data is stored for quick searches like this.
What can I search for?
You can search for anything really. There are various types of searches that are built out of the box like URL, Browser, IP, POST/GET data (form fields), etc. You can also do a text search across all data in the request/response and search on the UI data. You can use wildcards in your text search, a “*” is multiple characters and a “?” is a single character. If there are specific things you need to looks for like login, order number, etc. it is really easy to add (or simply use the text search). It is often suggested to add searches for login and order number to ensure a quicker search.
Active Session vs. Completed Sessions
When you search for sessions, one thing to be aware of is how active sessions and completed sessions differ on search capabilities. An active session is an active user on your site that still has not yet had 30 minutes of inactivity (time of inactivity is adjustable). You can also end sessions based on certain events like a session ending popup, etc. A completed session of course is a session that went beyond the time of inactivity.
An active session is holding ALL session data in short term memory (RAM). So, this is data that has not yet been indexed, so when you do a text search, you are looking across ALL data in the session. Often a search on active sessions will take some time, because it is looking across ALL active session data. When the session is moved from short term memory to long term memory, the data is indexed. This helps increase the search time on completed sessions. You should notice a considerable difference on completed session search even across multiple months of data. Now the thing to keep in mind here is that the HTML data is indexed but only data outside of the html tag. So, a line of html that contains the following will only have part of it indexed :
<span class=”errorMsg”>Error 1001: Insufficient Disk Space </span>
In this case the only parts that get indexed are – Error 1001 Insufficient Disk Space . Special characters are removed in the index as well as removed when you do a search.
Well what if you have some data on the page that you want indexed inside HTML tags? Well, there is a way to pull that data and push it into the appdata section in Tealeaf. This is done with a “reqset rule”, and anyone in Tealeaf consulting can walk you through doing this. Everything in the appdata section is indexed. So, I often work with clients that have metadata on the page that is used by web analytics, and other marketing tools. We push this data to the appdata section and we can search for it easily. Also, any JSON calls from the web page are normally inside script tags. So, pulling the JSON data into the appdata section ensures that you can search for this data anytime in the future.
Using Search with IBM Digital Analytics
Now, how do we use this powerful search with digital analytics? Well, this is where having IBM Digital Analytics definitely has its advantage. Doing a search across days, weeks or months returns a list of sessions. Much like the integration of segments from Digital Analytics to Tealeaf, the sessions from Tealeaf can be passed to the IBM Digital Analytics tools. This becomes a new segment inside IBM Digital Analytics to do further analysis. So, even though you may have missed something in your tagging, you can still pull back some interesting data based on a search within Tealeaf. This is extremely powerful and opens up your team to further analysis and opportunities. Sure we can do some of that analysis in Tealeaf, but this is where we talk about the power of Digital Analytics and the democratization of data. Tealeaf has a certain number of seats, but Digital Analytics is often spread across the entire organization and having these new segments in IBM Digital Analytics helps make the data more compelling. More on the democratization of data on part 3 of the series.
Want to drive real value from your Tealeaf installation? Leverage the CX Optimization Blitz and other offerings from the Tealeaf Analytics Services Team.
Contact Ryan Ekins to find out more. email@example.com
Follow me on Twitter: @solanalytics