How do YOU avoid "double dipping" on goal conversion values (micro + macro values = "goofy number")?
This is something I've always questioned.
- Micro Conversion A Value: $5.00
- Micro Conversion B Value: $10.00
- Micro Conversion C Value: $15.00
User completes all three. Value of user is now at $30.00
- Macro Conversion Value: $90.00
The micro conversions primarily exist to support the eventual occurrence of the macro conversion; they really have very little actual value on their own (or, significantly less value) unless a macro eventually occurs.
So now we have a $120.00 user (according to analytics) who has actually delivered $90.00 in actual business value so far.
- You depreciate the micro values further, i.e., try to figure out the ADDITIONAL value they generated to the business (i.e., on top of the macro conversion value). But this can be very hard to pin down. Other than contributing to the macro conversion value (their main purpose), you need to find out what other value that generated (maybe like they shared it on social? But then in this case, THAT (i.e., the share) becomes the micro conversion with value ascribed to it, and not the original conversion point (sign up for our email newsletter). So here I'm now actually removing value from some micro conversions. This kind of defeats the point of getting a feel for how valuable it is for someone to sign up for an email newsletter...
- You reduce the value of the macro conversion accordingly, to offset the value that's already been generated. Problem here is that users quite often don't NEED to complete micro conversions to complete the macro conversion, so you are left with bad/inconsistent data, since there will be varying cohorts/segments that completed a varying number of micro conversions.
All this said, what's inferred here is that I'd like the aggregate value that is shown in Google analytics to be a ballpark representation to how valuable the website actually is, and not just "some number"; if it's just a number then I certainly don't want a dollar sign before it and I can just create a custom "Value-o-meter" metric that I watch for relative changes.
Am I crazy to want this? Is it simpler than I've made it out to be?
Thanks. I'm sure some smart person(s) have figured out a good way to reconcile/approach this!