Highlighted

'Override Last-Touch Channel' - default setting causes completely incorrect marketing channel data

ryane_askew

15-10-2015

The Marketing Channels default setting is for the 'Override Last-Touch Channel' boxes to all be unchecked, but as I've learned, this actually causes a major problem with marketing channel conversion attribution. The correct default setting should be for all boxes to be checked. When none of the boxes are checked, the data does not collect based on touch points- in fact it completely disregards a first click/last click concept, and will attribute conversion credit purely on the ordering of channel process rules.

Here is an example of what happens when all the boxes are left unchecked:
Let's assume Paid Search is at the top of your processing rules, and Email is second. A user first visits the site from Paid Search ad, then a week later they visit from Email and convert. In this example, Paid Search will receive full credit for this conversion in the Last-Touch Channel report as well as the First-Touch Channel report.

Essentially, whatever is at the top of the processing rule list will trump all other channels below it (regardless of a user's last-touch channel interaction). Even if the user visits the 5 times from 5 different channels, if they ever visited the site via Paid Search, then Paid Search will receive full conversion credit.

In my opinion, this is a much needed change within the interface. There should be a clear warning of this potential problem as many organizations could unknowingly have this issue, and could be making major marketing decisions based on completely incorrect data.

Replies

Highlighted

Gigazelle

Employee

15-10-2015

I absolutely agree that leaving 'override last touch channel' unchecked on many channels will screw up the data really bad, but in my testing only the 'internal' and 'direct' channels have it unchecked by default. Customers were complaining these channels incorrectly stole credit, but all others should override last touch channel unless it is manually unchecked.

If you duplicated a report suite from one that had these disabled, it might have left them unchecked.

Highlighted

ecarpe15

15-10-2015

We currently have them checked to override except for direct and abandon session refresh.  

If you have this same setting, direct is a % of traffic to the channel.  (i.e. when breaking down paid search, one of the referrers will be direct (typed/bookmarked). Good to know when doing analysis on your direct channel, b/c it shorting the actual traffic from the channel and attributing to last touch channel that is set to override.  

This link my also clarify the usability of the feature: 

http://blogs.adobe.com/digitalmarketing/analytics/marketing-channels-overrides-who-gets-the-credit/

Highlighted

ryane_askew

15-10-2015

Hey Gigazelle- makes sense. My concern is really around the lack of notification about un-checking all the boxes- at the very least, Adobe documentation should be updated to reflect this potential problem.

I read through any documentation I could get my hands on, and even reached out to three different representatives from Adobe ClientCare about this issue, and none of these resources could determine the root cause of our channel data problem. The Adobe representatives even assured me leaving the boxes un-checked boxes couldn't be causing the issue and that it should behave the same as all the boxes being checked. Even though this was the feedback I received from Adobe from three different representatives, I still continued my own research and determined this was actually the cause of the problem.

I think there definitely needs to be more awareness about this potential issue whether it be a popup alert, better documentation, or education of the ClientCare representatives.

Highlighted

ryane_askew

15-10-2015

Hi Eric- thanks for the reply. I've actually read up a lot on marketing channels report before I posted about this issue, and that blog was very helpful in better understanding how this report works based on the configuration.

Highlighted

ecarpe15

15-10-2015

No problem, I have been researching a lot lately trying to understand how visits are credited.  Maybe you guys can help or offer advice on my question! How can you get an accurate visit count for the channels that are sandwiched between the 1st and last touch?

Visit 1 = Organic

Visit 2 = PPC

Visit 3 = Email 

When I pull my channels report, would the PPC show zero visits since it was overrode by the Email channel, or would it show one visit? Other metrics (orders, revenue, etc.) have a work around with linear and participation metrics that would spread the metric as I wished, but how does it work for visits? 

The fundamental problem we are having is in this example, the PPC channel manager is reporting a visit number that isn't in the same ball park as what Omniture is reporting.  

Highlighted

ryane_askew

15-10-2015

Hey Eric- visits are always counted no matter where it falls on a visitor's path. If your PPC channel manager (such as AdWords) shows 100 visits, Omniture should also show 100 visits. If these visit counts are different, signs point to a problem with your appended tagging or your rules for identifying those channels (such as the s_cid could be falling off in the URL parameters).

The concept of credit only pertains to the conversion (orders, revenue), so you will definitely see a discrepancy between your PPC channel manager and Omniture last click credit since your PPC channel manager does not "know" the true last click of a user like Omniture does.

Hope that helps.

Highlighted

ecarpe15

15-10-2015

That is what I assumed as well, but I chatted with client care yesterday and they said differently.  [img]Client Care Chat.png[/img]

Highlighted

ryane_askew

15-10-2015

Ah, I see now. This may help explain-

http://microsite.omniture.com/t2/help/en_US/mchannel/index.html#First_and_Last_Touch

"Every success event instance will have exactly one First Touch channel and exactly one Last Touch channel. This means that if you add up a given metric column for any success event, it will always exactly equal the total for the same time period. Non-success event metrics, such as visits and visitors, will not match up 1 to 1, as multiple channels can fire in the same visit."

This is probably the discrepancy you are seeing. Also, if you have channels set to override, the "last touch" visit will be taken away from the true last channel and given back to the second to last channel. This would mean that the channel being overwritten will have a lower number of orders and visits overall.

Locutus243

13-09-2017

Hiya guys,

Coming a little bit late to the party (about two yrs late)...but this seemed the perfect place to ask this question as it relates so closely to what I'm trying to do.

I've unchecked Direct and Session Refresh channels so they DO NOT over-ride the other channels. But the checkboxes for all our channels are ticked so that they WILL over-ride. But this causes me two problems:

1. Our sign-in is on a separate domain, so customers could come from site from a PPC campaign, but then when they sign in then it Adobe thinks they have come from regdomain.com. The problem being, that we have sister sites of which regdomain is one of them, so I actually want to track customers who come from Managed: Sister Sites, but NOT customers who have come from another source and then log in.

2. Our PPC numbers in Last Touch Marketing Channel don't add up to what we see in Ad-Words because a customer comes from a PPC campaign, thus setting their last touch channel to Managed:PPC. If that customer comes again that day Directly then because Direct is not set to over-ride then it increases the visits against Managed: PPC.

Any ideas/suggestions?

Thanks

Mark