Building "% of engaged visit" metric - validation help, please!

[jlatin]

I'm creating a custom calculated metric to better understanding engagement activity based on page visits.

To achieve this, I've set up a custom metric as follows:

Engaged user rate = (link click, download, form start, video view) / total visits 

Questions: 

1. Does this look accurate? Could I be missing anything?

2. How do I set the maximum percent to 100% in the metric builder so I'm not seeing these 100%+ figures at the aggregate level? (I set the upper % limit in conditional values to 100% and the lower limit to 0%) - what am I overlooking? 

[Jennifer_Dungan · Accepted solution]

So the first thing I notice in your calculated metric:

Engaged user rate = (link click, download, form start, video view) / total visits 

every visit will have multiple links, downloads, form starts and video views within a visit... so the % will generally always be above 100% based on this...

Are you trying to get something more like "Visits with Engagement / Total Visits" - to get the % of engaged visits?

If so, then you just have to make a slight adjustment to your calculated metric...

First, make a VISIT level segment:

VISIT

    link click exists
    OR
    download exists
    OR
    form start exists
    OR
    video view exists

This will get you the Visits with any of these engagement factors, for reference I am calling this Segment "Visits with Engagement Actions"

Now, in you calculated metric you can use:

Segment: Visits with Engagement Actions

    Visits

/

Visits

---

This should now get you the % of engaged Visits

If you are trying to get engaged users

Then you can do a similar segment using VISITOR scope (or even use the same VISIT scope segment, as Visitor who have ever had a visit with the scenario would still work), and in the calculated metric use:

Segment: Visitors with Engagement Actions  (or even Visits with Engagement Actions)

    Unique Visitors

/

Unique Visitors

---

In these examples, you are dividing based on the same scope... not mixing many raw actions against a much smaller denominator....

[jlatin]

This solves my issues, brilliant! 

Thank you, thank you! 

[Jennifer_Dungan]

You're welcome  

Sometimes it just takes a second pair of eyes