Foundations Matter: Aligning UX & Future Personalization Activity

September 5, 2017 | TBG Strategy/UX

Let’s talk about the intersection of page design on digital properties and their future flexibility and suitability for experience personalization. As discussed in previous blog posts, true personalization activity adds a digital “Z-axis” to the well-known “X-axis” of page layout and the “Y-axis” of browsing experience, in addition to the “leap across the browsing-experience tree” with search and links generated by metadata. (These spatial metaphors are a bit misleading but probably indispensable metaphors to consider regarding the organization of information, and in any case, the digital world already has many other cases of layering and swapping in play in contemporary UX.)

In many cases, personalization strategies are retrofitted onto existing sites, landing pages and apps; and they inherit whatever limitations exist there in terms of the kinds of elements that are available for personalization and their relationships to the rest of the experience. Conversely, our clients often ask us, if one is going to redesign something, what principles can one pay attention to that will best enable personalization strategy going forward?

Six Helpful Principles

Each case is different, but we have identified a series of six basic principles that are helpful, informing design and UX decisions when the goal is to have relatively rich approaches to personalization:

1. The Need for Target Elements, Or Places to Put New Target Elements

To the extent you want to personalize a digital experience, pages or views need to contain non-crucial elements that can be swapped out; the number is in direct proportion to the number and kind of personalization efforts you are likely to engage with in the future. For instance, if you want to personalize a homepage or sectional landing page within a site, but the page in question has an extremely minimal number of elements, it will limit the number and kind of personalization efforts that can happen in the future. It is also possible to insert new elements with personalization in some cases, but then those elements must still naturally work within the page experience—another version of the same consideration. Since the number of items on a page or of a certain type is a crucial consideration in relation to the perceptual aspects of usability and quality UX, inserting elements that aren’t well-tuned to the page is a non-starter.

For this reason, elements like featured topical content boxes, 3rd rail features, featured content in “mega-flyover” navigation and “super-footers,” featured content “slivers” in long scrolling pages, and the like, all play an important role in providing swappable elements that can be taken advantage of in the course of personalization efforts. 

As a result, personalization can directly impact page layout, putting a premium of having at least a few extra elements over minimal layout choices, or having thought through where extra pieces can be inserted. This is very clearly a case of “what’s good for design is not good for personalization,” in the sense a giant, highly-tuned, brand-building design with no swappable features or topical content boxes is likely going to frustrate attempts for downstream personalization. Figuring out the right balance of brand impact, clarity, usability, and swappable elements is key.


2. Personalized Elements Must Be Dispensable

Since the elements that are to be personalized are going to change, they can’t be elements necessary for the meaning and/or usability of the page. This issue requires assessment on many levels, for instance, in some cases a key organizational positioning message may not be able to go away for any users, or an element which is crucial to how users manage to transact can’t be replaced with something that doesn’t perform that function. Therefore, for targeted experiences to be visible, personalization generally needs to militate in the direction of having a tier of visible elements on pages that can go away without screwing up the most crucial baseline aspects of site experience. Looking at layout elements on a per-page basis through this lens is crucial. Again, it is always possible to insert new elements rather than replace elements, but in finely tuned user experiences, that is not always possible or desirable.


3. Priority of Personalized Elements

If personalized elements are going to be high impact and shift behavior, they need to be seen; it makes little sense to put a lot of effort into managing rules, generating content, and tracking results if the personalized elements are buried very low on pages or in obscure locations that don’t get much attention. Further, in some cases, personalized elements, by their associated strategy, must be part of the first impression (as in the case of personalized positioning, varying by visitor location, for instance). Thus, the items mentioned above might need to appear “above the fold” so to speak or at least high in the experience. Also, it is worth noting that since many visitors do not enter a site through its homepage, the considerations of personalization need to be thought of in terms of personalization throughout the site, for visitors who enter through Google’s back door. 


4. Some Swapped Elements Generate Greater Art Direction & Content Challenges

You need to think of the operational impact of what a personalization program’s content and imagery needs are—and if they are feasible. Some kinds of content or imagery are hard to create in large volumes or numbers of permutations, and create high risk situations from a design or photo-editing perspective. For example, large, iconic “hero” images (as on the top of a homepage) are in many ways an interesting target for personalization, but they are suggestive of the challenges of personalizing around “high design” elements. Coming up with five very strong text/images that work together (and with the site) for a rotating hero image is sometimes a challenge; now imagine adding 10 more text/image combinations to meet 10 personalized cases, and you’ll get some sense how using a hero image as the subject of personalization can stretch the bounds of art direction. It can be very hard to find 15 images that work together in different contexts, are all high impact, and all fit the feel of the rest of the user experience; in this sense, personalizing high design elements can generate design—and associated operational—challenges. The same is true for other high design elements, like elements which require individual illustrations, or for elements of content that are hard to write, like audience-specific positioning statements. Here, the underlying principle is that different kinds of elements chosen for personalization represent different levels of operational or outcome quality risk.


5. (Probably) Avoid Personalizing Structural Site Elements

When personalization makes structural changes to navigation or layout, it may impact usability. Users react well to personalized changes within narrow bounds, if they understand that more relevant content is being shown to them, or if the functionality presented is high value to them. They react poorly when the appearance of their favorite site or app changes in ways they can’t understand, and feel true cognitive dissonance when “that button they relied on to get to the next thing isn’t where it used to be.” For that reason, we must critically consider how users will react when elements like navigation, site structure, or major dimensions of page layout are changed by personalization. Users can feel angry and upset if they don’t like the site changes and if they don’t understand why they are happening.

The rules of thumb are:

  • the change needs to make sense to the user,
  • provide the user with something positive, and
  • include contextual cues that “not everything is changing.”

For this reason, much of personalization avoids these types of major structural changes. Therefore, base layouts and experience must be strong, and it isn’t generally a good idea to expect personalization to fix problems in those areas for specific user groups.

A few scenarios exist where personalization of structural elements makes sense, for instance when a site has well-defined audiences with radically non-overlapping needs (a site that caters to zookeepers and nurses, for instance). In those instances, if personalization changes the underlying structure and navigation of the site (which in any case is hard to do in many platforms, for complex sites), then there always needs to be a way to “opt out”— a way to undo it if it greatly impacts the experience or how the user can interact with the site. That, or at least to opt in. 

6. Atomic Design Strongly Supports These Principles

This topic is most likely the subject for a future blog post. Atomic Design—the shift to a regime of more modular and interchangeable design elements, from thinking of “pages” as gestalts—fits the reality of personalization, related content, variable page layouts, federated elements, multi-channel orchestration, and many other emerging digital realities. However, users experience pages as gestalts, not as collections of isolated elements. Modern design is born out of this fruitful contradiction, which in turn interacts with all of the considerations above. However, culturally, a turn to atomic design solves many operational issues and orients the UX and design teams towards solutions that statistically are more likely to line up with the needs of personalization.

Personalization Isn’t Just Flipping a Switch!

Personalization isn’t just a switch you flip! Many strategic nuances intersect with UX decisions and have strong downstream consequences, along with, of course, the great benefits they bring. The choice of what to include for personalizing the UX, how granular to get with personalizing elements (another subject for another post), and how to think about the overall interaction of elements in the user journey is important to make in a thoughtful way, especially if you want to create something sustainable and integrated. The principles we’ve articulated are a sound starting point to orient to the subject, but of course, each project and approach must be thoughtfully worked out on a case-by-case basis.

About the Author

person's silhouette with maze pattern
TBG Strategy/UX

TBG’s Strategy/UX team features resident virtuosos of digital business consulting, usability, information architecture, & CMS implementation & user testing. When you meet them, it’s impossible not to immediately catch that the digital world is their oyster.

Leave A Reply

comments powered by Disqus