Help me understand our form conversion metric when broken down by last touch channel. Please explain in details | Community
Skip to main content
Level 2
July 9, 2026
Question

Help me understand our form conversion metric when broken down by last touch channel. Please explain in details

  • July 9, 2026
  • 2 replies
  • 25 views

Hello the amazing Adobe Community,

 

I am looking to understand a metrics.

  • We have form conversion set as Segmented “Visit Container” with event## to fire when user submit a form
  • Then create a calculate metrics using that Visit Container Segment + the (Visit) metric is our Form Conversion

 

We want to understand why when we have last touch channel broken down and when we sum the form conversion for each of the channel it doesn’t add up to the “Deduplicate total within the workspace” example using dummy numbers. Can someone explain what is causing this? We understand that visits and sum should have slight difference of approx. 5% but for our form conversion calculated metrics its >15% difference between workspace number and total number. Explain in technical term is also welcome. I want to understand Adobe logic.

 

Channel Visits Conversions
Adobe Total 370,123 9,723
SEO 108,351 2,052
SEM 125,824 3,878
Direct 158,885 5,819
SUM 393,060 11,749

2 replies

parthGupta
Level 2
July 11, 2026

Hi ​@ericnguyen021 

This isn't a bug or a wrong configuration on your end. It's a well-documented consequence of how "Most Recent" (Last Touch) eVar allocation works combined with visit-level breakdown reporting.

Your Marketing Channel eVar is likely set to Expiration: Visit / Allocation: Most Recent. This means the channel value can be overwritten mid-visit every time a new channel-qualifying event fires (a new campaign click, paid ad click, referral, etc.). The value credited to your conversion is whatever the eVar shows at the moment the form is submitted — but a breakdown table shows a visit under every channel value it held during that visit, not just the final one.

Test: Isolate visits with more than one channel value
Build a segment:

Container: Visit

Hit container 1: Marketing Channel = SEO
Hit container 2: Marketing Channel = SEM
Hit container 3: event## exists
All joined with AND

This finds visits where SEO and SEM both occurred (on different hits) alongside a form conversion, within the same visit.


Check: Apply the segment and look at the Visits count. If it's a meaningful chunk of your 9,723 total conversions, that's direct proof channel-switching mid-visit is driving the gap.


To fully validate: Repeat for other channel pairs (SEO↔Direct, SEM↔Direct) and sum the Visits across all — that total should roughly match your >15% discrepancy.

Jennifer_Dungan
Community Advisor and Adobe Champion
Community Advisor and Adobe Champion
July 11, 2026

Yes, like ​@parthGupta said…. Marketing Channels are “last touch” meaning the last set value persists; but visits can touch multiple Channels, and when broken down, each visit would have to be counted at each channel, but in total, those visits are de-duplicated. Especially if you are looking at a different Attribution Model...

 

I will try to illustrate this with a single visit:

 

  • (User performs a Google Search)
  • Page A
    • Marketing Channel set to “SEO”
    • Marketing Channel Detail set to “Google”
    • Marketing Channel Instance is triggered
    • Page View is triggered
    • Visit is triggered
  • Page B
    • Marketing Channel maintains value “SEO”
    • Marketing Channel Detail maintains value “Google”
    • Marketing Channel Instance is NOT triggered (because no channel is actually “set”)
    • Page View is triggered
  • (User gets a Marketing Email and clicks on a link - this is the same Visit)
  • Page C
    • Marketing Channel set to “Email”
    • Marketing Channel Detail set to “Campaign A”
    • Marketing Channel Instance is triggered
    • Page View is triggered
  • (User decides to look up info about other promotions, performs a search and clicks on an ad)
  • Page D
    • Marketing Channel set to “SEM”
    • Marketing Channel Detail set to “Campaign B”
    • Marketing Channel Instance is triggered
    • Page View is triggered
  • Page E
    • Marketing Channel maintains value “SEM”
    • Marketing Channel Detail maintains value ‘Google”
    • Marketing Channel Instance is NOT triggered (because no channel is actually “set”)
    • Page View is triggered
  • Conversion
    • Marketing Channel maintains value “SEM”
    • Marketing Channel Detail maintains value ‘Google”
    • Marketing Channel Instance is NOT triggered (because no channel is actually “set”)
    • Order is triggered
  • Page F
    • Marketing Channel maintains value “SEM”
    • Marketing Channel Detail maintains value ‘Google”
    • Marketing Channel Instance is NOT triggered (because no channel is actually “set”)
    • Page View is triggered
  • (User Checks their Social Media, and clicks on a link)
    • Marketing Channel set to “Social Network”
    • Marketing Channel Detail set to “FB”
    • Marketing Channel Instance is triggered
    • Page View is triggered

 

Now, let’s look at how Adobe would show this:

  Page Views Visits Orders

Orders

Participation|Visit

Marketing Channel Instance
Marketing Channel 7 1 1 1 4
  SEM 3 1 1 1 1
  SEO 2 1 0 1 1
  Email 1 1 0 1 1
  Social Network 1 1 0 0 1

 

 

Four different channels were touched in the same visit… when they are broken out, they all need to show “1” as they were part of a Visit; but the total Visits is still just 1 (not 3)… this is the de-duplication at play.

 

SEM was the last channel to be set before the Order (and the value persisted) so it gets the credit for the order; however, using Attribution IQ, and looking at the cross channel attribution for ALL Channels in the Visit, you can see that SEO and Email both are given credit as well (since they were touched before the order occurred; but the Social Network channel happed after, and therefore has no weight or attribution to the Order.

 

It’s possible that your Order might be looking at Participation, and not just standard “Last Touch”?