Create Pivot Tasks in Bionic Rules


Userlevel 7
Badge +5
Pivot tasks help you transform data into a polished, actionable dataset in Bionic Rules. In the 6-min. video below, Karl Rumelhart, Gainsight's Chief Product Officer, shares a quick demo on how to configure pivot tasks to transform event data. 



Let us know if these short DIY videos are helpful! We're hoping you might be inspired to record your own DIY video and share it here on the Community! To share your own video here, upload it to vimeo or youtube for example, and then add the link in your Community post.




12 replies

I can't wrap my head around when or how something this would be useful. Is anyone doing this and for what purpose?
Badge
I really love the hands-on demo. Can you give us some examples of what regular actions you would design based on this? I always feel like I'm doing it wrong if I'm trying to make a report out of a bionic rule and not making actions off of it. 
Userlevel 7
Badge +3
Yes I love how to videos so much more than reading docs! Both are necessary sure, but for us visual learners the videos are super helpful!



Are these going to be available on the regular support site too?
Userlevel 7
Badge +5
Thanks, Jeff! Yes, I'll be adding these videos into GS Go too. Have you had an opportunity to use pivot tasks in bionic rules? Can you share any real life use cases? 🙂
Userlevel 7
Badge +5
One possible use case is to conduct this pivot, and calculate a Daily Average User metric, and set a Call-to-Action to alert the CSM that John George has zero page views, all in one rule.
Userlevel 7
Badge +2
Use cases for Pivot tasks frequently come up on community. Here is one example where our Rules Engine expert Jitin came to the rescue recommending the Pivot function: [filter_by]=all&topic-reply-list[settings][reply_id]=19234716#reply_19234716]https://community.gainsight.com/gainsight/topics/concatenate-in-rules-or-reporting-engine?topic-repl...
Badge +4
Use cases for pivot tend to be when you have lists of things of different types and you want to look at them by type.  Here is one probably everyone can appreciate:



Suppose you have a list of Open Cases.  One of the columns is 'priority' with values like P1, P2 and so forth.  You can pivot this around and then see for each customer how many P1 cases, P2 cases, etc. 



To Scott's comment, there are then tons of potential actions.  You may want to set a health score based on how many cases there are of different priorities, or create a CTA or something like that. Of course, you could also write the data into an object which will then make it easy to report on the pivoted data. 
Badge +4
One reason I created this video is to show that anybody can do it!  (And you don't have to have fancy production values to be useful.)   Think of something that you know how to do in Gainsight and create a short video showing how.  There will *definitely* be people who benefit.    Who knows how to set up associated objects for CTAs?  Can someone review the color controls in the product?  How do you build a custom Activity type?    You know tons of stuff!  Please share. 
Userlevel 2
Badge
We use these pivot tasks not for CTAs (yet) but in setting some of the measures of the scorecard.  You can see how the number or logins (or in our case, we use number of touch points with the CSM) can easily be aggregated and pushed to the scorecard all in one rule.
Userlevel 7
Badge +6
This video did a great job at helping me figure out why Pivot tasks were getting me so confused! :-)



When Pivoting in other applications, you don't need to explicitly define each Pivot column. For example, I want to Pivot on a date field. My Fetch task is pulling in data based on Last 30 Days, so I need the Pivot task to dynamically identify which dates are present in the data, and make each date its own column. Is that possible?
Lila and Karl,

The video tutorial is an awesome idea. I am sure a lot of users are going to benefit from this.



Seth,

The option to explicitly define each Pivot column because it may not always be all unique values for which you need columns. It may be a subset of values. Also, there might be groups/aggregations for which you want to create columns. (Score greater than 5 as one column, score <5 as another)

For a use case where you want a column for each unique value with the same aggregation(eg. count), I agree that this can be made much simpler. Thank you for the feedback.

Pivot on a date field is not supported in Bionic Rules currently. Lets use this thread to list our other use cases, where Pivot on a date field might be required.
Userlevel 7
Badge +3
I just hope I never need to use it!

Reply