Adobe Target Archives - MiaProva

Tag: Adobe Target

Activity Conversions Use Case

This week, we had a client reach out to us with an interesting use case with one of their Activities in Adobe Target.

The Activity was a test in their cart targeted to visitors who were making their third visit to that location. Adobe Target profiles easily support targeting such an audience but the Analysts needed to only count purchases if they happened during the same session as that third visit.

This video highlights how you can use Adobe Target ‘events’ and data like Profile Attributes as Activity Goals and still have the data in Adobe Analytics via A4T.

Video Transcript

And screenshots of the “Primary Goal Configuration” to automate the flow of this data into Analytics:

Adobe Target Conversion Events in Adobe Analytics Workspace

Adobe Analytics Activity Conversions

MiaProva makes heavy use of Activity Conversions via Adobe Analytics and because of that, we have gotten quite a few questions about scenarios where it might not work given the timing of the clients side calls of Adobe Target and Adobe Analytics.

This post outlines a few test cases that we did to help our customers understand just how helpful this utility is and to encourage the use of this bonus success event from Adobe Target.

For those of you not familiar with Acstivity Conversions, think of it as a success metric or bonus metric that comes directly from Adobe Target and is fed into Adobe Analytics for analysis. The most common use of this bonus metric is using the Adobe Target VEC to track clicks on content. We also have many MiaProva customers that use Adobe Target Profiles combined with Target server calls (mboxes) as a metric as well. Since the Target Profile isn’t in Analytics (yet), this is a cool approach.

The Three Test cases

High level – Adobe Target and Adobe Analytics leverage both client-side and server-side components to bring you A4T. Adobe Target uses a server to server integration to port over Activity Impressions and Activity Conversions.  We validated this across three tests using an Adobe Target conversion metric and all three showed up in Analytics.  This includes a scenario where there is NO Analytics call after the Adobe Target success event.  

Adobe Target Conversion Events in Adobe Analytics Workspace

Test 1 – Adobe Target ‘clicked element’ success event via VEC with Adobe Analytics on next page

First test – A4T with an Adobe Target Clicked Element Conversion event (element was a link to another page on the same site with the same Analytics account in place – so there would be an Adobe Analytics page view call on that page).  Validated that click mbox call was taking place upon click but no analytics call until the second page loaded.

Findings:  Data from Target, Classic Analytics Reports, and Analytics Workspace is PERFECT aligned.  **In Analytics Workspace, you have to change the attribution model to “same touch” otherwise the data will be inflated BIG time.  

Unique VisitorsVisitsWorkplace Activity ImpressionsTarget ImpressionsAnalytics Reports Activity ImpressionWorkspace Activity ConversionsTarget ConversionsAnalytics Reports Activity Conversion
Activity Conversions in Adobe Workspace (Clicked Element Conversion)2628313131252525
Experience A1515171717141414
Experience B1113141414111111

Test 2 – Adobe Target conversion event ‘URL contains’ via VEC and FORM

Second test – Instead of using a Target Click Event, I used Target Conversion “URL contains X”.  This would guarantee an Analytics call would take place.  

Findings:  Data is PERFECT between Target, Analytics Reports, and Analytics Workspace.  

Unique VisitorsVisitsWorkplace Activity ImpressionsTarget ImpressionsAnalytics Reports Activity ImpressionWorkspace Activity ConversionsTarget ConversionsAnalytics Reports Activity Conversion
Activity Conversions in Adobe Workspace (URL Conversion)910292929666
Experience A66111111555
Experience B34181818111

Test 3 – Adobe Target VEC ‘clicked element’ success event with NO subsequent Adobe Analytics call


Third Test – I used a click event that DID NOT lead to a subsequent page load therefore, NO Adobe Analytics Call would take place since click happens after page render and first Analytics call took place.  

Findings:  Data is PERFECT between Target, Analytics Reports, and Analytics Workspace.  It is odd though that the Visit counts were slightly higher than Impressions though.  Conversion data HAS to be server to server and not client-side.  No Analytics call is needed for Target success event Conversion data to be seen in Analytics Workspace.

Unique VisitorsVisitsWorkplace Activity ImpressionsTarget ImpressionsAnalytics Reports Activity ImpressionWorkspace Activity ConversionsTarget ConversionsAnalytics Reports Activity Conversion
Activity Conversions in Adobe Workspace (No Analytics Click Event)6468666666555
Experience A4041424242000
Experience B2427242424555

Take away – Activity Conversions, by way of A4T, can be a great mechanism to get data into Adobe Analytics that normally wouldn’t be there. And because of the server-to-server integration between Adobe Target and Adobe Analytics, the Activity Conversion data WILL be in Analytics no matter if there is an Adobe Analytics call after the success event took place.

A4T – Activity Conversions

The Adobe Analytic’s Product team had a pretty significant product release last week that will be greatly appreciated by the A4T and the MiaProva communities. The issue regarding inflated values for Adobe Target Activity Impressions and more importantly, Adobe Target Activity Conversions has been resolved. This will significantly enhance the value A4T brings to Adobe Target users and help Adobe Target users see the power of Activity Conversions – an Adobe Target Metric in Adobe Analytics.

I suspect some background might help here. In this world of A4T, end-users of Adobe Target could go into Analytics Workspace and classic Analytics Reports and dig into their test results and mine for nuggets of gold. Below is an example of such a report – a Workspace freeform table with the Adobe Target tests being the dimension.

A4T - Adobe Target results in Analytics Workspace
Adobe Target results (A4T) in Adobe Analytics Workspace

Up until last week, the Activity Impression and more importantly, the Activity Conversion counts above were inflated in Workspace and via the API (how MiaProva works:). These values are correct in classic Adobe Analytics Reports and in Adobe Target as seen here:

Adobe Target Activity Conversions Adobe Analytics Reports
Adobe Target in classic Adobe Analytics Reports
Adobe Target A4T Reporting
Adobe Target reporting for an A4T Test

This issue hasn’t been the biggest of deals for A4T users because often end users of Adobe Target don’t configure a legitimate conversion event in Adobe Target or they select an Adobe Analytics metric instead as I have above with Trial Starts [E7]. This update though should help users see the value of using the Activity Conversion – or more specifically a bonus success event from Adobe Target.

Use the Activity Conversion – Adobe Target Metric

Yes, when you select Adobe Analytics as the Reporting Source as seen in the image below, you can simply select an Analytics event or metrics but don’t. Doing so is a waste because you get any and all metrics or events in Analytics Workspace or Classic reports automatically without selecting it here.

Here you are given an opportunity to include an Adobe Target success event which will be incredibly helpful when it comes time to dive into the results. With any A4T test (a test selected with Analytics as the reporting source is an A4T test), you get ONE Adobe Target metric to be part of your test results and that is what is called Activity Conversions in Adobe Analytics Workspace and Adobe Analytics classic reporting.

A4T - Adobe Target success event
Adobe Target metric configuration with A4T Activity

The Target metrics available are:

  • Viewed a page – very helpful when you want a quick way to track impressions of a URL or list of URLs. This can be tricky to configure in Analytics sometimes so helpful to have this here.

  • Viewed an mbox – this is perfect for any time you are using nested mbox or Adobe Target calls in an offer or test. I personally use this one quite a bit if I am evaluating the DOM via Target and need to track specific events or things where the test is running

  • Clicked an element – this one is HUGE and only an option for VEC Activities and not those that Form-based. This metric will allow end-users to quickly and easily track clicks on elements where the test is running. Often Analytics isn’t configured to track clicks on everything on the site and if it was, finding that metric in Analytics could be challenging.

For the “Viewed an mbox” and the “Viewed a page” options above, you have the option of applying Adobe Target Audience Targeting as well. This means any shared Experience Cloud Audiences from AAM or your Adobe Target profiles can be put to work to refine any metric as needed.

A4T Audience Targeting Activity Conversions
Audience Targeting in A4T Target Metric

Activity Conversions and MiaProva

Within MiaProva, we do a pretty cool dance behind the scenes between Adobe Target I/O and Adobe Analytics I/O. We support any and all Activity Types from Adobe Target and we heart those A4T tests as well including the Activity Conversions metric!

In the MiaProva admin section, MiaProva Test Managers standardize reporting for Adobe Target Activites and A4T tests. We’ve defaulted (users can modify though) to prioritizing “Activity Conversions” for A4T metrics to help users take advantage of this bonus metrics.

MiaProva A4T Activity Conversions

Want more, do more, get more from Adobe Target – augment your 1st Party ID

Raising functionality awareness with the MiaProva Chrome Extension

Our Chrome Extension (CE) was publicly launched just over three months ago and the adoption and usage has been incredible.  If you use Adobe Target, in any capacity, it is definitely worth checking out and this post highlights some of the features and capabilities if you want to learn more about it.    At a high level, the CE will make you a much better user of Target and will allow you to get considerably more value from the Adobe Target platform.

This blog post stems directly from one of the strategic features in the CE: visibility into an organization’s 1st Party ID and all the data that is associated with that ID.  Several people have asked me about this 1st Party ID highlighted so prominently within the CE ID tab, so I figured it would be helpful to writeup exactly what it is and how users can leverage it to enhance or augment their testing and personalization efforts.

2nd Party

Adobe Target does an amazing job at providing organizations with the ability to run tests with many different test types.  That ability exists because Adobe is providing organizations with a Visitor ID (mboxPCID).  That Visitor ID from Adobe has data associated with it such as the type or browser, device, time of day, URL of pages seen, etc…  Any and all targeting rules that you use to setup your Activities in Adobe Target uses data that is mapped to that 2nd Party ID (Adobe Target’s ID).  Also, important to note that any and all Adobe Target profile attributes and Activity participation is also mapped directly and associated with this 2nd Party ID.

This 2nd Party ID is what the vast majority of Adobe Target users solely rely on. That is 100% totally fine and cool as significant value can be realized with that setup. Adobe Target though, has something that allows users to build on that 2nd Party ID world which enables organizations to center their testing program around their ID – their 1st Party ID vs. just relying on the 2nd Party ID.

Adobe isn’t limiting organizations to operate solely within Adobe’s walled garden but rather, allows the Adobe garden to be rolled up into an organization’s garden (1st Party), giving the organizations garden a higher priority! MiaProva’s CE provides visibility into data from both of these gardens for end users to consume, modify, and even augment all in one nicely laid out location.

1st Party

Facebook (probably not a good example given press as of late), Google, Amazon, Finance, Telco’s, Travel, Retailers with loyalty programs, and countless other business types have their own Visitor ID. I can’t think of a single website or digital property that doesn’t have a Visitor ID. Some optimization teams though, might not have visibility into exactly what it is and how to get it but that is another topic. Trust that there is an ID managed by your company on the site and that there is data associated with it. Heck – many organizations have several Visitors IDs that they manage.

When you sign in or authenticate with whatever login ID that you have (email address, cell phone number, etc….) you are then becoming known to the organization which allows an ID to be created and data to be associated with it. That data can and should be part of your optimization and personalization efforts. Retailers have all that incredible loyalty data, Finance has credit scores, transaction history, Telco’s has phone upgrade data and countless other items, and Publishers has consumer behavior. The opportunities are endless.

That same ID, an organizations 1st Party ID – can be easily shared to Adobe Target. In fact, if you have an ID like this on your digital properties (you all do:), you should absolutely share this with Adobe Target as it costs you nothing and opens the door to a whole new level of strategic testing. Once this 1st Party ID is shared with Adobe Target that opens the door for the organization to shift the mindset of their personalization efforts around their data and their data lakes versus having to rely solely on data other parties collect or don’t collect. This move alone is symbolic for many reasons as it allows the organization to unify around the broader internal understanding of who a visitor is.

Adobe Target and Visitor ID’s

MiaProva Chrome Extension ID Tab

The image to the right is from the ID tab of our CE. It is this tab and the content within it that has lead to so many inquiries around what this 1st Party ID is why it is blank (not many Adobe Target users have this in place so we purposely left it blank if we don’t see a value). In this image the value is 5628258 as that is what we are telling Adobe Target it is.

The Adobe Target Visitor ID value in the image is the ID that Adobe sets and manages. This is a 2nd Party ID from your organizations perspective and this ID is for Adobe Target.

The Experience Cloud Visitor ID (formally the Marketing Cloud ID and still that in some documentation) is another 2nd Party ID that Adobe provides and this ID is shared with both Analytics and Target as well as Audience Manager. This ID is what enables A4T if your organization uses that. And coincidentally, this ID, by way of Audience Manager, opens your testing program to 3rd Party ID’s and all the data associated with those data providers! Fun world to live in when you have a 1st Party, 2nd Party, and 3rd Party ID’s working together

mbox3rdPartyId

To enjoy the world of both 1st Party and 2nd Party data, a very minor addition or tweak to your current implementation must take place. This tweak works with all versions of Target, whether you use mbox.js or at.js. If you leverage tag management tools for your management of Adobe Target, this should take just a few minutes.

Essentially, we need to add a new parameter to your Adobe Target call. This parameter has to called ‘mbox3rdPartyId’ and you would pass to it your 1st Party ID value. Below is how it be done with at.js using the targetPageParams function (has to be called before at.js). You would replace “<%:1stPartyID%>” with your server side variable that represents your 1st Party ID.

function targetPageParams() {
               return 'mbox3rdPartyId=<%:1stPartyID%>;';
            }

In the image below, I highlight what it looks like by way of the MiaProva CE. In the Activities tab, users can click on an active Activity and this allows them to view the mbox parameters being passed to Adobe Target.

mbox3rdPartyID
MiaProva Chrome Extension mbox parameters (mbox3rdPartyId)

Once this special parameter is passed to Adobe Target, the 1st Party ID is then married to the Adobe Target 2nd Party ID.

Awesomeness

When this marriage has taken place, organizations should look at the various data points available and establish some strategies around how best to make use of it. Because of how easy it is technically to marry these ID’s (literally 5 minutes of work:), organizations typically learn about the capability of using 1st Party ID data and immediately put it in place immediately before laying out strategies.

The strategy work can typically be a big process in and of itself and will likely involve many meetings with various groups to outline ways to make use of this data with a testing mindset. Lots of opportunity and if one wants to just jump into the deep end immediately to see what 1st Party ID data is impacting consumer behavior, one might want to leverage Automated Personalization as an Activity Type. That Activity type makes use of 1st and 2nd Party ID data when making decisioning. I often see 1st Party ID data bubble its way up to the top and surprise end users as to what data is impacting model based decisioning.

Once strategies have been laid out we have to move on how best to get this rich data over to Target. This data can be communicated client side or server side. Where the data lives often determines the communication path. Personally, I find using the APIs to be the easiest as they are straight-forward and you can do up to 500,000 rows in a single upload. Here is the documentation on this approach.

Considerations

When this ID is in place, it takes the higher priority from Adobe. For one thing, this ID enables cross device test coordination. This is HUGE. So, for example, I get Experience C when at my work computer in the office. When I come home and use my iPad, if this same ‘mbox3rdPartyId’ is seen again, I am automatically put back into what I was originally given. This is a massive component and will help with your current testing program and results even if you have no 1st Party data.

Because of this cross device coordination of test or Activity experiences, QA’ing can be tough for end users because Adobe Target will automatically remember what Activity and Experiences were associated with that ID. End users can clear their cookies until they are blue in the face and Adobe will automatically show the same Experience. Keep this in mind as you setup tests. Having extra 1st Party ID’s will be very helpful in this instance.

MiaProva 2018

Like you, we love data…

As we are closing out the year here at MiaProva, we took a step back to examine some numbers and trends over the last year. Like you, we love data and thought we should share just a few of the interesting tidbits here.

MiaProva Application

22,988 Adobe Target Activities are being managed within MiaProva. This is significant because that means that 22,988 tests or Adobe Target Activities can be automatically included into their Program Overview dashboard for real time visibility into key organizational metrics. These Activities are also actively monitored and are ready for consumption into an organizations easily consumable Knowledge base.

MiaProva’s Program Overview

22,988 tests is a HUGE number and we are quite proud of it. This wouldn’t be possible without the automation we’ve created by taping into Adobe I/O. Here is a breakdown of the Activities by type:

A/B Activities: 15,916

XT or Experience Targeted Activities: 6,920

ABT or Automated personalization, Auto-Allocate, Recommendations, MVT, etc…: 152

MiaProva works seamlessly with any test or Activity type used in Adobe Target. It is going to be interesting to see if the ratio of advanced test types like Automated Personalization or Auto-Allocate changes in the months and year ahead. We expect it to considerably and our Chrome Extension should help facilitate that a bit.

MiaProva Chrome Extension

We launched our free Chrome Extension (CE) for all Adobe Target users back in September and the adoption and use has been great. It has been used 252,090 times since the launch up until today, 12/31/2018.

Activities

One of the key components of our CE is visibility into Adobe Target Activities and the interaction of those Activities with one another. In the image below, we can see that on this page (our home page), we have 4 active Activities that we are in at this location and as a Visitor, we are associated with another Activity in a different location.

MiaProva's Chrome Extension and visibility into Adobe Target Activities
MiaProva’s Chrome Extension’s Activities Tab.

Having 4 tests running on the same page is definitely not ideal for many reasons including attribution reporting and code conflict. Many organizations run into this challenge especially with more democratized programs and that is why we included this as the primary tab of our CE.

Below is a breakdown of the number of Adobe Target Activities by page over the last three months. As you can see, while the majority of tests per page is one, there are a significant number of test locations with many more than one running. Two test locations were found to have a whopping 15 Activities running!!

Adobe Target tests by page impressions.

Profiles

Adobe Target profiles are arguably the most powerful component to the Adobe Target platform. We wanted organizations to have visibility into what profile attributes were and what values were associated with their Adobe Target Visitor ID (mbox PCID).

The image below represents the profile attributes associated to my Adobe Target ID. Adobe Target users can use this on their site to familiarize what is associated with them, they can edit values, and then can even add new values. Pretty powerful for many reasons but especially helpful for debugging.

MiaProva's Chrome Extension Profile Tab - visibility into your Adobe Target Profile.
MiaProva’s Chrome Extension Profile Tab

While it is not a true representation of how sophisticated a testing program is, the number of profile values does highlight an organizations familiarity with this component of Adobe Target.

Over the last three months, we have found the number of profile attributes varies considerably. The average number of profile attributes was 62.43. The most common number was 14 and the highest number of profile attributes seen was 323!

2019

There is a lot to look forward to in 2019. Optimization and Test Management solutions have proven to be an essential part of any organizations testing program. We are quite proud of our ability to help our clients this year: realize the impact they are making into organizational metrics, incorporate efficiencies and visibility around the entire testing process, and see that not a single test goes to waste with our automatic knowledge base construction.

We are full steam ahead on this mission and wish you all a happy and optimized 2019!