Many conversations at optimization events this year are focused on personalization, and delivering “the right content to the right person at the right time.” I am surprised by how many companies still find themselves mired in mere isolated testing in high-value locations such as home or conversion pages, and evaluating results at a super-high “new vs. return visitor” level. Inevitably, many will admit that they haven’t quite gotten to targeting and personalization . . . yet.
Truthfully, there is conversion and revenue lift to be seen by targeting the best-performing content at this super-high, very basic audience or segmentation level. When employing this basic level of targeting, we must ask ourselves what we are missing in terms of critical subsegments of the test population and their preferences. How are visitors performing on subsequent pages in the funnel after a test, when they’re dumped into a more “vanilla” and less personalized experience? What revenue, conversion lift, and brand loyalty is being lost when these critical subsegments are overlooked in the testing and targeting of relevant, personalized content?
In the nascent years of optimization, there were many valid excuses for why a testing and optimization program could not evolve beyond isolated high-value testing into more advanced targeting and personalization. First, there could be limited creative resources. Test variations required heavy lifting by a design team to build, and several experience variations were too costly and time-consuming to create.
Second, utilizing only basic a/b testing tools or functions limited the ability to test sequences, or multiple pages operating together in one test. This inhibited the ability to determine a test’s influence further down in the experience and to provide sequential personalization for key segments. Limited success metrics in the tool and insufficient compound audience or granular segment filters (think “Mac User Mothers Driving Minivans from Montana”) made it difficult to determine distinct visitor preferences, which could very well differ from the test population as a whole.
Third, there was an inability to target this audience appropriately in real time. Whether manually or through automated means, the technology didn’t exist to dynamically update content based on test results. Fourth, scalability was an issue. People were unable to run many concurrent tests while maintaining accuracy or to ensure that these tests did not adversely influence the test results of another test running within a particular region of the site. Finally, there was a barrier to entry in terms of understanding how the testing tool worked and implementing best practices for building successful and high-value tests.
With recent advancements in Adobe Target, companies can now scale and mature more quickly from beginning incidental testing to more advanced testing and personalization. At a recent Adobe Summit tech lab, I witnessed optimization managers who were new to Target beating the teacher through the program’s learning process and executing on tests right away. In fact, Target’s guided, visual interface and workflow has reduced training from three days to a couple of hours.
The new approach reduces requirements for creative resources in terms of experience variations using advanced WYSIWYG (“what you see is what you get”) editing capabilities within Target’s Visual Experience Composer. This provides the nontechnical marketer with easy editing functionality including the ability to move, insert, and rearrange any content within any container on the page, including text, imagery, and overall design. All of this with built-in safeguards that protect against breaking the page code.
The tool enables advanced testing scenarios, such as linking tests in emails, mobile and social apps, and display ads, to onsite or sequential page experiences in order to accurately determine the best journey for valued segments across channels and touch points. Automated behavioral targeting assists in these scenarios as well, when you’re unsure of how to target an individual. This requires the modeling system to learn and self-optimize over time based on the best-performing content and the most predictive variables at a particular touch point or page. Not only does this personalize your content to a specific, known segment, it also offers segment discovery opportunities. Having several algorithms to choose from, and to test them against your own homegrown algorithm, is what we’re enabling with our automated personalization functionality.
Recent data synchronization between the Analytics and Target solutions allows for applying any success metric or audience segment defined in analytics to test reports. This creates unlimited drill-down capability in reports to determine the most relevant and profitable opportunities for targeting or to better define strategy. It also lets you measure and target winning content in real time, either manually or self-optimizing. Multilevel targeting, built-in prioritization, and collision alerts allow you to scale to more concurrent tests in a location with accuracy safeguards and an enhanced ability to direct test traffic for clearer, more effective results.
Regarding ease of entry, you can invite any team member or department across your organization to be part of Adobe Marketing Cloud and allow them to build test designs that you can prioritize with the intuitive interface. Target’s guided workflow educates novice testers on what they should be thinking about in terms of test design and measuring to effectively qualify their results. Roles and permissions let you manage more people, as well as oversee and prioritize tests based on business needs and cost-benefit analyses. An additional ease factor is the one-step single line of code implementation, which lets you implement and get up and running immediately, or use a tag management solution for an even faster start.
The new Adobe Target breaks many of the perceived barriers to optimization maturity, allowing marketing teams to expand their testing programs to profitable personalization across all of their touch points, quickly outperforming and leaving the basic, baseline conversion lift of incidental testing in the dust.