When I am multiplying Visit x Conversion rate there -.2% discrepancy against Orders?
Does anyone know why this would be happening?
Views
Replies
Total Likes
Do you have anything else happening in your table or panel? Any Segments that might be causing some minor issues? Are you running this on a Virtual Report suite that might be using a segment that is causing a discrepancy?
When I do this in my data, it lines up exactly. If you can't show your data here publicly, would you be able to share a small sample directly with me in private messaging? So I can see your setup and maybe be able to spot something that I can't from your description?
Views
Replies
Total Likes
A discrepancy as small as .2% (either positive or negative) sounds like it might be a rounding error. When you multiply visits by conversions you could end up with a number that isn't a whole number (ex, 10 visits and a conversion of 23% would equal 2.3 orders). But in actual data, you can't have a partial order, either someone places an order or they don't, so something like 2.3 orders wouldn't be possible, it would have to be 2 orders or 3 orders. When you get into big numbers and multiple different time periods, it becomes more possible that if your rounding numbers there will be a slight discrepancy between an estimate and an actual number.
Views
Replies
Total Likes
Except that the default "conversion rate" metric (this is an out of the box metric provided by Adobe) is literally "Orders / Visits"... so if you take that metric and multiply by Visits, it should match orders exactly...
I was also thinking it might be a rounding error on a custom conversion rate.... but Visits * (Orders / Visits) should work out right back to match the number of Orders, there shouldn't be any rounding error...
I've checked my own data, as well as the Adobe sandbox data... when I create the multiplication, every row ends in .00 (there is no discrepancy).. also, when you make the multiplication, it converts "Conversion Rate" back to the basic division...
So this shouldn't result in a rounding error... it's not taking a rounded version and then multiplying, it's doing all the division and multiplication in this metric...