Simplicity, Not So Simple: Embracing Deep Gradual Engagement

Making an experience simple doesn’t have to be about reducing complexity for the users; it’s about attempting to stay within what the users are initially capable of managing, and providing them increasing options as their familiarity increases. The user experience field has matured enough to recognize what those user states are across applications, devices, and even environments, but is just scratching the surface when it comes to taking advantage of the users’ evolution within applications and websites.

Classic Persona Modeling Is Only a First Step

Persona development has enabled UX to take a more human-centered approach to building a website and meeting the largest percentage of users’ needs. While this works for simple sites with little depth, a more complex site requires a more complex approach to identifying and facilitating users’ needs. Personas are a necessary step in our process, but can now branch out to match users’ needs, as users become more advanced in their interactions. For a very long time, technology limited presentation to anonymous users and, perhaps, alternate states with different and additional content once a user voluntarily authenticated with a username, tying the session to a specific identity.

In recent years, the Internet changed; it became more complex by degrees. The hardware and infrastructure became more sophisticated. Sites today have access to more information, in real time, about the browsing session, without violating the privacy of individual users. Sites have discrete access to social infrastructures when explicitly provided by the user. Sites can receive data from third-party sources, by user permission, and respond to information that never has to be accessed or stored by the site providing the interface.

Instead of the simple model of “click, request, respond,” there is now a complex series of options and possibilities; a tangled and complicated journey with every click, previously limited by users’ ability to use the technology. Conversely, as designers, we tend to limit users by what we offer them. There are still some who assume that sites should be built to the least common denominator throughout, instead of just initially, and in treating every user as if they have a sixth grade reading level, an introductory understanding of computers, and the attention span of a gnat.

The users of the Internet have become more familiar with standard web browsing techniques and user interfaces. Some of them started to trust the websites they used. And some of them got really good at using them. So good that they are often able to learn faster ways of using websites and inform designers how to better design. Those users are ready for additional functionality and have the current capacity to learn new methods associated with using it.

Parallels to Progressive Disclosure

This idea isn’t new; designers practice progressive disclosure, based on interface restrictions and/or amount of data, relative to the needs of the user. Edward Tufte’s visualization theories have allowed designers to explore this concept, through the lens of partitioning information, presenting it in pieces that create a new whole to assist users in better understanding interfaces. Typically, this disclosure state is made based on a single set of user requirements and rarely revisited for those users who are engaged with the interface at deeper levels. By taking the feedback gathered from that initially identified user state, an interface is engineered that presents just the right amount of information within each set of choices, being careful not to exceed users’ ability to comfortably make a decision.

Louis Rosenfeld expanded on this concept with a blog post in which he compared users’ capacitive levels to an onion. On the surface, the initial disclosures hold true. As the layers of the onion are peeled, the number of users diminishes, but they become more skilled, engaged, and trusting of the interfaces they’re interacting with, along different segments of their journey. This model allows for a quick, simple overview of how you can segment users into more complex personas that allow you to respond to multiple sets of needs.

The fear of a user struggling with the way forward can create an unintended over-simplification of interfaces. Opportunities are missed by remaining on the outer peel of the onion. It’s easy to identify a user who has been to a website seven times in a month. Once a user has been to a website multiple times, it becomes likely that more of your interface has become resident in their long-term memory. You can then offer additional content, helpful to a more advanced user; you can provide shorter ways of navigating sets of functionality, previously complex to the user’s novice state. Recognizing and emphasizing the facets of your site that will have become a part of long-term memory will allow you to avoid potential information overload. provides a good example of this with their latest website. Upon initially visiting the site, you’re presented with essential navigation and promotional material. However, after browsing through several products, a new widget appears in the bottom-left of the pages, granting you quick, easier access to previously viewed products. They do this based on a cookie, without logging you in. While the feature itself is simple, the revelation of that feature matches the user state in that the user is continuing to shop, or is a return visitor, and thus, a more engaged user than the first-time visitor who has no need for an empty “previously viewed” container.

Examples of Moving from Simple to Complex

Math is taught to varying degrees from grade school through post-education. The core principles need to be taught and become rote before advanced, complex math can be introduced. Children must learn addition before subtraction; those both before multiplication, then algebra, geometry, and trigonometry, on through calculus and beyond. Algebra is possible because the tenets of more basic math have become embedded in long-term memory.

Video game developers have noticed this and have shaped their games accordingly, with increasingly complex stories and parallel learning tracks. With early video games, such as Super Mario Brothers, the system of control and interface was fairly simple, as was the game. Once you learned the controls for the first level, you could viably skip directly to the last level and still succeed, provided you could keep up with the number of factors, but the interface remained the same. In more recent years, complex games such as Spore and World of Warcraft require the player to learn and store interface controls within long-term memory, to allow the constant introduction of new abilities, techniques, and controls. A first-level player of World of Warcraft suddenly exposed to an end-game level interface would not be well-equipped for success, having missed exposure to over one hundred required in-game lessons.

The Case for Advanced Features

Users carry different levels of commitment, training, and trust, based on where they are in a given process. They may have goals; they may have a required task. They may not. Regardless of their intent, their level of familiarity extends to how they proceed with matching what they perceive as needs. With greater familiarity comes trust in the system to do what they expect it to when they perform actions (see Figure 1). With that trust and familiarity comes the desire for additional, advanced features. Requirements for experiences should be able to match this cycle of needs.

Table 1. Parallels Between Trust Achieved and Time Commitment
Users can be plotted along the parallels between trust gained and time committed.
Asking for Information (Gradual Engagement) Providing Information (Progressive Disclosure)
1. Initialized Passive. One size fits all, choice theory observed. What can be assumed for all users? Categorical. What do you need to do (A), or (B), or (C), or (D)
2. Re-Visited Search.
Direct relation to request. What are you looking for?
Having selected (B) before, are you interested in topic (1), or (2), or (3), on in going back to Category?
3. Engaged Personally Identifiable.
First Hane, Last Name, Email, or Phone.
How can I help you, Jane Doe?
Having reviewed the components of (B), (1), (2), (3), here is additional information about (B) as (4) and (5).
4. Re-Engaged Patterns in site engagement.
Confirmed through multiple contact points.
How was your experience? What can we do for you today?
Multiple Levels.
Here is the most recent information about (B>, (6) and (7), and also information within (C) that closely relates.
5. Committed Trust.
Personal planning, storage of information.
We can offer you (x), based on what you’ve been working on. Are you interested?
Deep Levels.
Having explored (A), (B), and (C);, here is the information you’ve found most relevant, and any pertinent supporting information.

Experienced users, engaged at a deep level, are looking for a quicker, more efficient way to do introductory tasks while adding the ability to do additional work to their stable of rote tasks. By modeling out the levels at which users engage, and by identifying how their needs change or expand as they become more engaged, you can take a design approach that can gradually become more complex, allowing for more advanced features. Appropriately complex interfaces allow for richer, greater functionality and expanded capabilities.

When to Use Deep Gradual Engagement

The ideas of gradual engagement, progressive disclosure, and simplification are inextricably linked. The role of the UX practitioner is to understand the engagement levels of the users and know when it’s appropriate to reveal functionality to them.

Much of this comes from work that UX practitioners already do: Who are your users? What does the business want them to do? How will the users respond along the way? This is traditionally and generally all that needs to be defined for simple projects. But not every project is simple.

There’s a greater level of granularity available regarding users that can be written and designed to. The key is to understand how many levels of information you have for your own functionality and how engaged you will be able to keep your users. Exposing information or functionality as users need it can be daunting and dangerous if done incorrectly. Remembering to still provide intuitive access to unrevealed features is paramount.

Luke Wroblewski illustrated gradual engagement as a sign-up strategy, clearly articulating the gains Twitter benefitted from once they adopted it. By not leaving users to their own devices once they had signed up, and by stepping them through the process of how to successfully use the application, there was a 29 percent increase in use conversion.

Once you’ve determined that your system is complex enough that it requires a deep engagement strategy, the creation of a User States Grid can help you begin to understand your users and the opportunities you have (see Figure 2). The key information to collect for a User States Grid includes the following:

  • User Levels. These can start as arbitrary labels; you will get a better idea for them as you complete your research.
  • Trust Level. This is the level of trust the user has in the brand, site, or system.
  • Technical Recognition. Factors that are part of the site or system that allow you to associate a binary identifier with the particular user level.
  • User Data. The type of data that is known and/or stored about the user.
  • Exposed Functionality. The associated functionality with a given user level.
  • Conversion Impetus. The factor that gives the user a reason to move to the next engagement level.
Table 2. An Example of a Deep Engagement Strategy
The grid is a starting-point example of how users can be plotted to determine how they best align to a deep engagement strategy for your website or system. Additional user levels can be identified and added, as appropriate, to a given project
Trust Level Technical Recognition User Data Exposed Functionality Conversion Impetus
1. Initiatized Low None None Primary Navigation Cues Discovering Product
2. Re-Visited Low-Medium Cookies Browse History via Cookies Additional History Features Purchasing Product
3. Engaged Medium Registered and First-Time Authentication First Name, Last Name, Email Address, Password (encrypted) Recommended Products Based on History Home Improvement Purchase Information Access
4. Re-Engaged Medium-High Return Authentication Encrypted Purchase Info, Additional PII Info Recommended Products Based on Purchases Home Improvement Project Planning Access
5. Committed High Authentication after Multiple Interactions Personalized Home Information, Family Information Recommended Products Based on Home Information N/A

The level of complexity regarding your gradual engagement strategy will be dependent on how detailed you and your organization or your client want it to be. It can be an initial engagement, to better teach users the value of your site, or it can be a prolonged strategy, to continue the education and benefits available to your users for their entire lifetime use of your site.

At Lowe’s, we’ve adopted this strategy for how we release our MyLowe’s home management platform. Initially, we were able to plant a set of tools in the hands of our users and provide them a meaningful system to react to. We’ve come to realize, however, that by walking them through various steps of the journey, responding to the amount of information they’re willing to provide, and giving them tools and interactions commensurate to their level of understanding of our system, we provide them with a more engaging experience that we can continue throughout the entire lifecycle of the user being a Lowe’s customer. We can grow with them.

As an example, when we initially released our “Save Item” feature, we provided users the ability to save directly into any one of multiple, pre-defined favorite lists or pre-defined rooms. Additionally, the users could actually create a new room or a new list from within this area. We discovered that many users were confused when they initially interacted with this feature, lacking the context of use provided by the tools they were referencing. As a result, we reoriented our strategy to provide only the most basic level of save functionality, later increasing it once the user had interacted with the referenced tools. While this is one small example, this is the level of scrutiny we apply to each feature as we generate experiences that evolve with our users.

Putting aside conventional notions of what the web was, and moving into a new era of developing interfaces appropriate for multiple levels of engagement will help strengthen the very foundation of the anticipated experience. Our users have complex and varying needs. We have the technology to serve those needs. Until now, we haven’t taken advantage of it. By identifying who the users are, throughout the lifetime of their engagement with our interface, we are able to better serve their needs. We can evolve not just how we present information to our users, but enhance our ability to efficiently serve advanced users. Doing this without hindering our ability to bring novices into our site enables us to produce the most positive experience possible for all of our users.UX

Comments are closed.