One of my favorite chapters from renowned mathematician Jacob Bronowski’s magnum opus The Ascent of Man is titled “Knowledge or Certainty.” In the chapter, the author uses a variety of scientific devices and methods to attempt to define the human face. After lengthy measurements, tests, and results, the face is then put into a human context, which alters any scientific result that had been objectively measured previously. You learn that the human being measured is a prisoner of war, which places all previous measurements in a different context from that previously perceived, creating a different personalized perspective and meaning from what any objective device is capable of recording. Bronowski then begs that math and science, that measurement itself, should never be devoid of humanity or conducted outside of a human context, which helps to interpret and accurately report its results. He quotes Oliver Cromwell in relation to the scientific method, “I beseech you, in the bowels of Christ, think it possible you may be mistaken”

This example, and the possibility of being “mistaken,” is the fundamental reason for iterative testing. No longer can we rely solely on pure data or interpretive hypotheses and hunches alone; accurate analysis requires a combination of the two when determining the right experience to deliver to a valuable audience segment of our digital visitors. As we’ve covered in previous blogs and recent announcements at Adobe Summit, having unified Adobe Analytics/Target/Marketing Cloud data in a real-time aggregate profile view of your customer is essential. This multifaceted understanding of the customer allows for greater clarity and accuracy in measuring and uncovering the most predictive profile variables and opportunities for content personalization and maximizing return on investment (ROI) relative to your primary business goals.

This is why relying on the basic out-of-the-box segmentation in a point optimization solution for an overall view of your visitor population is not enough. You need to uncover the most profitable granular opportunities, through manual or automated means, based on all of the available historical and real-time data on your visitors. This allows you to gain an accurate view of where valuable opportunities for personalization exist, and where they do not (equally important), to effectively define your strategy. Not only does this affect your immediate returns but it also improves growth and scalability within your testing and targeting to increase effective interpretation of your results.

This means digging deeper than the basic segmentation of new vs. return visitors, or behavioral targeting, to determine if things like time of day, recency or frequency of visits, or geography play a valuable role in determining if one experience is more relevant and profitable than another. Often going a layer deeper and customizing compound segments with the most predictive variables can promote larger increases in engagement and exponential returns when targeting your visitors across your digital touch points.

Granular data and custom audience segmentation capabilities are not the only factors that impact accurate, effective testing and optimization. Having detailed success metrics in your test reports is also extremely valuable. For instance, let’s say I’m only tracking click-throughs from an email campaign test to the landing page, and I don’t look at additional key performance indicators (KPIs) down the funnel (e.g., interaction with landing page offers, recommendations, or even page depth/consumption/time on site). By using single metrics or simplistic out-of-the-box tools I might interpret the data incorrectly, or misinterpret the human visitor impact of my test down the funnel. The email test variation might have been successful in terms of generating basic click-through, but caused unforeseen friction later in the customer experience if these metrics were overlooked, or if the solution was unable to capture them because of a lack of customization.

Adobe Target allows you to dynamically apply any and all success metrics to analyze the relative impact of test variations on visitor behavior. This is where it’s immensely valuable to have synchronized Adobe Analytics success metrics to apply to results, even retroactively, as enabled by Adobe Marketing Cloud’s unified definition and application of audience segments and success metrics between Analytics and Target. This allows for further qualification of results with a broader, more detailed view of performance provided by the full context of Analytics data. Target also provides the ability to apply revenue and custom scoring to success metrics and page values relative to consumption. This provides the revenue impact and detailed reporting needed to empower you to immediately show the positive impact of your testing and optimization efforts relative to the values and revenue streams that matter most to your executives and company. Target is the deeply customizable, “plug-and-play” solution that can immediately make you a revenue driver and optimization hero within your organization, especially as you re-platform or redesign your site, launch new marketing campaigns or applications, or look to implement a cross-channel campaign solution.

For example, one of our clients is a large financial institution whose main conversion goals are new accounts, new investments, and cross-sells. The company began with testing and targeting based on geography (what banking opportunities were available by region) and basic behavioral targeting (what category of investments a visitor was viewing). Naturally they saw some initial impressive returns, which is always valuable in fueling your program. Given these initial gains, many businesses might have deemed this a success and stopped there. This financial institution, however, used Target’s APIs to regularly import updated online and offline branch profile data as well as third-party data such as credit score, to create a scoring table for prequalification of different offers based on a customer’s level of investment with the bank and other qualification criteria. The company was then able to match the customer to the most relevant offer based on a matrix of possible offers to display. By doing this, they saw lift in hundreds of percentage points in click-throughs from the homepage. Again, they could have deemed this a success and called it a day. However, they were also able to construct and apply custom success metrics, such as customer lifetime value, to determine the larger impact of delivering more personalized offers in terms of their customers’ longevity and happiness as well as the exponential returns the bank was experiencing through this relevant personalized engagement.

A cross-channel view of all of these success metrics, or custom combinations therein, along with the expanded view of analytics-enhanced reporting, gives you the best set of reporting data on which to apply your human interpretation, allowing you to be “less mistaken” in your judgment as well as to accurately ascribe the business value and revenue lift that your detailed data-driven decisions have generated.