Insights

Home » Data-Driven Hyper-Personalization via A4T (ADOBE ANALYTICS + ADOBE TARGET)

Data-Driven Hyper-Personalization via A4T (ADOBE ANALYTICS + ADOBE TARGET)

Reading Time: 23 minutes

Personalization across channels and by touch point, is an imperative for the organization that has made a commitment to provide a personalized experience for their customer, i.e., true one-to-one marketing.

At least that’s been the line of thought for the past 7-to-8 years or so.

Seemingly flying in the face of this axiom, was a report published late last year by Gartner which stated, “80% of Marketers Will Abandon Personalization Efforts by 2025 due to lack of ROI, the perils of customer data management or both. In fact, 27% of marketers believe data is the key obstacle to personalization — revealing their weaknesses in data collection, integration and protection.”

This doesn’t seem completely out of step today given the death of the 3rd party cookie and the uncertainty as to how we’ll transact as a society going forward. At the same time this assertion buttresses the value proposition Adobe has brought to market in the server-to-server integration of Adobe Analytics and Adobe Target, or A4T. This powerful solution set drives ROI via the optimization of data and content, and effectively addresses privacy issues.

Let’s begin by discussing the value of the A4T integration at a macro-level in the context of the Gartner report, and then transition into the tactical utilization of Adobe Target and Adobe Analytics to drive personalization across your website, and through Adobe Target’s API to Single Page Applications (SPAs) and mobile channels, and on non-browser based IoT devices such as connected TV, smart devices, in-store digital screens, digital consoles, set-top boxes, and digital kiosks in stores and airports.

Adobe Target Personalization via the Internet of Things

Lack of ROI

*Before we proceed it should be noted that data variance, reporting issues, and ROI attenuation are present in all software integrations where there are disparate methods in data collection. This is not an Adobe issue – this a technology issue that challenges some of the largest brands in the world. The logical argument presented henceforth is that by integrating Adobe Analytics with Adobe Target reporting issues and ROI attenuation due to data variance will be eliminated.

The Adobe A4T integration enables marketers to optimize key metrics in Adobe Target that were created in Adobe Analytics, i.e., revenue, orders, applications submitted, calculated metrics – virtually any metric that your team has identified as a measure of true ROI. These metrics are set during the creation of your Adobe Target Activities and Experiences. The utilization of Adobe Analytics metrics in Adobe Target is a key component behind how our clients at Innover have been able to realize a 40%+ improvement in site conversion KPIs.

A primary impediment of personalization ROI in the past, via web analytics and personalization software platforms, was the significant variance in data coming from the personalization platform into the analytics platform – and Adobe was not immune to this issue. According to Adobe, prior to the A4T integration there had been a 30% variance, on average, in reporting due to different counting methodologies across applications. This led to hours, if not weeks, of analysis to identify the right opportunities for personalization, and to accurately identify and report on the ROI of a company’s personalization efforts. Adobe Target’s server-to-server integration with Adobe Analytics (A4T) has remedied the 30% variance reporting issue, allowing customers to truly understand the insights from the personalized experiences they are delivering and, moreover, optimize the ROI of their personalization campaigns.

Furthermore, this server-to-server integration means optimization opportunities can be surfaced in Adobe Analytics and immediately acted upon with confidence within Adobe Target. In addition, a complete view of ROI impact is automatically provided within both Adobe Target and Adobe Analytics. A marketer is able to accurately measure the entire customer journey, from a click on a display ad, to an ad viewed on a connected TV, to receiving a personalized offer in your mobile app.

In addition to the real-time insights afforded by the A4T integration, marketers are able to ask retroactive “What if?” questions for further analysis of Adobe Target campaigns. Any Adobe Analytics metric, dimension, or segment can be looked at retroactively for additional insights. You can select Adobe Analytics as the default reporting source during the Adobe Target activity creation process. Then within Adobe Analytics you can select any success metric or audience segment defined in Adobe Analytics and retroactively apply it to your reporting for extensive filtering and drill-down analysis of your optimization results. In other words you are not relegated to insights and optimization based solely on the metrics you use to set up your Adobe Target activities, you are also able to drill into metrics and segments that were not used directly in the Adobe Target activity to ask a question like, “How would this audience or metric have performed in my Adobe Target campaign?”

Let’s drill down on the ROI implications of the data variance, and why the variance exists outside of the A4T integration.

The problem arises if the data coming into your Adobe Analytics instance from your Adobe Target instance is inaccurate, and your marketing team is utilizing this inaccurate data to create audience segments in Adobe Analytics, which are then being used to create activities in Adobe Target. It is a closed-loop cycle of failure from the start.

The activation of intelligent insights from Adobe Analytics in Adobe Target is of paramount concern for marketers. Marketing teams cannot afford a lack of data integrity – certainly not a discrepancy in data of 30% or more.

But why would there be a variance in data between Adobe products? Why the need for A4T?

In 2009 Adobe acquired the Salt Lake City based company, Omniture. Omniture created Adobe Target and Adobe Analytics (known by different names at the time). It was a questionable acquisition at that time – a $1.8B investment for analytics and A/B testing software by the company that makes Photoshop? Those products formed the foundation of Adobe’s Digital Experience business, which is now on a run rate of being a $3B business in FY20.

Although Omniture created both Analytics and Target, the way in which the platforms collect data, the amount of data collected, and the manner in which that data is actioned on, is much different. Moreover, because of the different implementation techniques and codes, Target and Analytics may collect data at different points in time. In a perfect design these actions would fire simultaneously, but in reality, because of the disparate implementation design and practices, the actions fire differently. Again, according to Adobe these disparate data collection approaches can lead to hours, if not weeks, of analysis to identify the right opportunities for personalization and to accurately identify and report on the ROI of a company’s personalization efforts.

The way in which Adobe Analytics and Adobe Target counts visitors is different as well.

Adobe Target and Adobe Analytics record unique visitors and session duration differently. In Adobe Target, a new visitor is someone who doesn’t have the Adobe Target cookie. They may be visiting the site for the first time, they may have recently cleared their cookies, the Mbox may be new, or they haven’t been to the site in over *2-weeks (*there are special considerations on cookie duration). In Adobe Analytics, that same cookie scenario is true, however, the cookie can persist for up to 2-years. There have been recent changes in Safari and Chrome that influence data collection and cookie duration, but that is outside the scope of this white paper.

Additionally, Adobe Analytics counts any visitor coming to a page, however, Adobe Target only counts a visitor if they qualify for a given activity. To further confound the integrity of our data collection, it is possible that the same person might qualify for activities whose destination is the same location. And although the visitor would see different content on the page, they would be counted twice in Adobe Target where in Adobe Analytics they would only be counted once.

At times you may need Adobe Target to leverage 3rd party cookies (via at.js 1.x). For example, if you own multiple websites that are live on different domains and you want to track the user journey across those websites, you can use third-party cookies by leveraging cross-domain tracking. However, this is a moot issue if using at.js 2.x since neither third-party cookies nor cross-domain tracking are supported in at.js 2.0.0. If you don’t know why utilizing 3rd party cookies to track the user journey across websites is problematic, you can read an excellent article here by our friends at Hubspot. By utilizing Adobe Analytics in place of Adobe Target to track the customer journey, the 3rd party cookie issue is resolved.

Customer Data Management seems intractable

Adobe Target enables the integration of Adobe Analytics data, CRM data, and 3rd party data, for the purpose of optimizing Adobe Target success metrics. Some examples of 3rd party data could include geo data via a visitor’s IP address purchased from a weather service, or household data purchased from Experian Marketing Data. In fact, Adobe Target enables brands to connect visitor interactions across channels with a unified, progressive profile that spans anonymous visitor to known customer. Interactions encompass both engagements online (site, app, email consoles, devices and screens) and offline (call center, in-store/branch, POS).

An example of utilizing 3rd party data to personalize the visitor experience is to deliver a hero image that correlates to the weather of where a visitor is located. On the webpage below the visitor is coming to your website from Park City, Utah, so Adobe Target knows to serve the visitor a banner that reflects their location. The call to the 3rd party weather service is made via the Adobe Target API. What’s amazing about Adobe Target is that you can scale your creativity to include a myriad of data points that encompass all aspects of behavioral, psychographic, demographic, and transactional data. The personalization opportunities are endless. This tremendous scale is enabled by Adobe Target’s main personalization algorithm (Random Forest machine learningused in both Automated Personalization and Auto-Target.

Personalize your webpages for unique visitor experiences

Privacy is a concern

Adobe Target allows for very personalized messaging to visitors based upon very granular (non PII) data. However, in light of the General Data Protection Regulation (GDPR) and the California Consumer Privacy Act, Adobe Target offers protections to ensure the privacy of visitors to your website.

According to Adobe, “one strategy to ensure privacy is opt-in functionality support via Adobe Launch to help support your consent management strategy. Opt-in functionality lets customers control how and when the Adobe Target tag is fired. There is also an option via Launch to pre-approve the Adobe Target tag.” Privacy and Adobe Launch are outside the scope of this white paper, however, should you need more information the team at Innover can support that inquiry.

Now Let’s Personalize

So, presuming you are able to resolve the lack of ROI, the perils of customer data management, and privacy concerns – personalization should again be the focal point of your digital GTM strategy.

But before we go deeper into personalization, let’s define what an Activity & Experience in Adobe Target are.

An Activity in Adobe Target lets you personalize content to specific audiences and test page designs.

For example, you might design an activity that tests two different landing pages, one page might offer visitors from Santa Cruz a $20 discount on surfboards, while another page might offer visitors from Park City a 10% discount on skis. The activity determines the conditions that control when each of these landing pages appear, and the metrics that determine which page is more successful. The activity is configured to start and end when specific conditions are met, such as between specific dates, or to start when the activity is approved and to end when it is deactivated.

When designing an activity, you should plan carefully. Determine when the activity will start and how long it will last. Then, list your offers and assign an audience segment to each one.

An Experience determines which content displays when a visitor meets the audience criteria for an activity.

An experience can be an offer, image, text, button, video, or combination of these various elements on a page or a set of pages that perhaps form a purchase funnel or some other logical sequence of pages. It can also be the response of a voice assistant, a customer service script, or even a personalized flavor from a drink machine. You test or personalize experiences in a Target activity.

Now that we are clear on what an Experience and an Activity are in Target, we are ready to jump in and start creating Target marketing campaigns – but before we do, there is one very important component of a successful personalization model that needs to be addressed. The marketing, eCommerce, and analytics teams are all vital to the success of your organization’s personalization efforts, and should therefore be inextricably linked to the mutual success of the campaigns.

Synchronization of Analytics Insights, e-Commerce, & Marketing Teams

A4T brings together the eCommerce, analytics and marketing teams. Oftentimes the individuals whose remit it is to derive insights from analytics data may not be closely aligned with the individuals whose remit it is to increase site conversion or improve CX. By providing visibility into Adobe Target activity success metrics to the analytics teams, and sharing Adobe Analytics conversion data with the marketing and eCommerce teams, this additional context for each team can serve as the catalyst they need to augment the success of their initiatives. Better yet, by creating one Digital Intelligence Platform through the combination of Adobe Analytics and Adobe Target, all three teams may feel compelled to work more closely with each other, which is certainly a best practice for high-performing B2B and B2C websites.

While bringing these teams together to work as a symbiotic e-Business entity is a best practice, it is also understood that they may still have a preference to work within the tools they feel most comfortable with. By integrating Adobe Analytics and Adobe Target these teams can continue to work in these separate technologies, all the while knowing the cost and conversion data they see in Adobe Analytics and Adobe Target is originating from the same conversion tags and the same report suites.

Let’s move to how we execute in Adobe Target and Adobe Analytics

It is time to get tactical and demonstrate where the rubber meets the road. Below is a screenshot of data (conversion and traffic metrics) that have been collected in Adobe Analytics being presented within the Adobe Target Activity Report. The number of metrics you can collect in Adobe Analytics isn’t infinite, but it is certainly more than what is available in Adobe Target (the number of eVars you are allotted in your Analytics instance determines the number of metrics you can capture). Within Adobe Target you are relegated to 5 success metrics – Conversion, Revenue, Page Views, Time on Site, and Custom Scoring. To view the Target success metrics combined with the Analytics conversion metrics in the context of Target Activities & Experiences, you would simply click on View in Analytics.

A4T Report in Adobe Target

When you click View in Analytics you are able to access the Analytics for Target (A4T) panel where you can analyze your Target activities and experiences in Analysis Workspace. The A4T panel also enables you to see Lift & Confidence for up to 3 success metrics. To access the A4T Panel, navigate to a report suite with A4T components enabled. Then, click the panel icon on the far left and drag the Analytics for Target panel into your Analysis Workspace Project.

A4T Report in Adobe Analytics

After we drag the A4T panel over, we can look to measure the efficacy of each experience – using Experience A as the control experience and choosing 3 success metrics to measure against. Visitors are the denominator in our measurement, where the 3 success metrics are the numerator to understand conversion rate by experience.

In the screenshot below we see a breakout of experiences and how each experience performed in terms of Conversion Rate, Lift, and Confidence for one of our three success metrics – Analytics Challenge Registration. Again, Experience A is our Control Experience, and we are looking to see how the other experiences have performed versus the Control Experience. The Lift metric compares the conversion rate range by variant for each experience against the control experience, and the Confidence metric tells us the likelihood that the results would be duplicated if the test were run again.

A4T Advanced Visualization Options

An additional advantage of bringing Target data into Analytics is the ability to leverage reports and visualizations not available in Adobe Target – such as Fallout and Flow visualizations.

Fallout Visualization

Fallout reports show where visitors left (fell out) and continued through (fell through) a predefined sequence of pages. A Fallout visualization displays conversion and fallout rates between each step or touchpoint in a sequence.

As you can see below you are able to assess fallout touchpoints (All Visits, Cart Additions, & Orders) by experience. You can compare each experience side-by-side in a Fallout visualization to see how fallout differs downstream.

Flow Visualization

The Flow visualization helps you understand the exact journeys your customers are taking on your website or your app. For example if you’d like to know what pages your customers are viewing and in the exact order they are viewing them you can use the Flow visualization. Below you can see where all of your customers are coming from to get to the home page and where they are going to from the home page. This can all be analyzed in each experience.

You might be asking, “Now that we’ve integrated Adobe Target with Adobe Analytics via the A4T Panel in Analysis Workspace, what can we do besides collect really granular insights and report out our findings?” Well, one thing you can do is create custom Audience Segments from all of the data you’ve captured in Analytics, i.e., 3rd party data, CRM data, marketing automation data, and of course data from Adobe Target.

Let’s imagine you run digital marketing for Veterans Insurance – an online banking and insurance company. Your customers are very loyal. When they buy a financial product or open an account, they stay for years – sometimes a lifetime. The problem you have is moving them into new financial products and/or having them open new accounts. Therefore, you decide that you are going to run a campaign across several channels to target existing customers that own one or more products – but that do not own car insurance.

It turns out that your campaign is a success and you achieve record traffic and application submissions. The problem is you also have a record number of incomplete applications. So, you decide that you want to retarget those visitors across channels, including your website. Given that objective, you’ll create an audience segment in Adobe Analytics and use this new audience segment to target those members that come back to your site with an experience and message that will compel them to finish and submit their application.

In the steps below we will demonstrate how we build high-value audience segments from the data we collect via the A4T integration, and how we then utilize those segments to optimize our Target activities.

Step 1

In Adobe Analytics click on Components > Segments

Step 2

Click on the + Add icon

Step 3

Add your title and description, and then add your Components (metrics, dimensions, segments, date range). In the scenario below you decided that you wanted to add people that have clicked on an email, and then came to the site and started an application, but never finished that application – within the last 7 days.

After you’ve added your Components you then want to click on Publish this segment to the Experience Cloud.

Step 4

Now you see that your new Audience Segment titled Auto insurance retarget has been published in Adobe Analytics.

Step 5

We head over to Adobe Target and click on Audiences. You can see that the audience we created in Adobe Analytics is now available in Adobe Target.

Step 6

You can now create an activity to retarget your customers who did not finish their application for a new car insurance policy on your website. We won’t go into the Activity Types, however, you can find information on Activity Types in this Adobe interactive PDF.

Step 7

In the final step before you create your experiences, you can choose the audience you created in Analytics –Auto insurance retarget – to apply to your A/B Test (Activity).

So, why doesn’t every company that leverages Adobe Analytics as its “source of truth” use Adobe Target for personalization?

There are too many reasons to cover in earnest, so I will share a few of the reasons that tend to surface the most.

We use a homegrown solution for personalization today. There are advantages to using a personalization solution you created from the ground up. It checks all of your boxes, the UI is customized, the integration into front end and back end systems are designed by your team, and the solution is built to your business requirements. However, there are a myriad of problems with a homegrown solution. It was probably much more expensive to build than first planned, and the current cost to iterate, let alone maintain product viability, is likely intractable. This used to be a prudent strategy when software wasn’t hosted in the cloud and automated updates weren’t churned out at the pace they are today. Additionally, if your e-Business is looking to scale your company will never be able to keep pace. According to Forrester’s Digital Intelligence Playbook, “Traditional cycles of analyzing customer behavior and then making manual changes to website or mobile apps to see if they affect engagement KPIs are too slow; they can’t keep up with the changing propensities of customers to respond to experiences.” Firms need to invest in solutions that leverage real-time and closed-loop processes that constantly learn from every customer interaction and apply new insights to optimize KPIs.

We use (Optimizely, Google Optimize360, Monetate, etc.) for personalization. A few years ago using a “Frakenstack” was an acceptable business practice; however, that thought process has since changed given the pace at which SaaS companies have acquired point solutions and have integrated these solutions within their Cloud ecosystem. How Apple, Walmart, Macy’s, USAA, and Nike function today, as compared to how they utilized marketing technology even 5-years-ago, is night and day. Companies are placing their bets on integrated SaaS solutions in the Cloud. That might be Google, Salesforce, Oracle, Adobe, etc. The fact is that these forward looking companies understand the multiplicative value and economies of scale that these SaaS companies offer, and one component of that multiplicative value is the exponential impact of utilizing two (or more) disparate software solutions to optimize the same business objective in a way that could not be realized on the merit of their individual output. Again, perhaps the paramount reason to use Adobe Analytics + Adobe Target is to ensure data integrity, which will ultimately improve the optimization of your personalization efforts. According to Adobe, “Variances of 15-20% are normal, even with similar data sets. Systems that count differently can result in much higher data variances, as much as 35-50%. In some cases, variances can be even higher.”

In the end your company needs to be able work with marketing technology solutions that supports your business objectives, and the utilization of Adobe Target might not meet all of your business requirements; however, if you are looking for a closed-loop cycle of insights, testing, and optimization, then Adobe Analytics + Adobe Target is the far and away the industry leader, and Adobe has the consensus of the industry analysts and success stories to prove it.

The Forrester Wave™: Digital Intelligence Platforms

The Forrester Wave™: Experience Optimization Platforms

If interested in a demo of the A4T solution, or if you’d like to discuss in greater detail, please reach out to our team. We’d love to help.


  • Adam Higdon  |  December 14, 2020   |  
  •  

Leave a Reply

Your email address will not be published. Required fields are marked *