When it comes to creating digital experiences, personalization and customization are terms that seem so similar and yet are vastly different in strategy and execution. Choosing the user experience (UX) option that makes the most sense depends on nuances of context and engineering — and even your company’s culture.
Customization: Doing things your way
Customization is about the user taking control. Take music streaming service Spotify, for example. Listeners build their own experiences through playlists. Popular task management applications like Trello are also customized experiences. Users create boards, choose the look and feel they like best, and add in their own workflows. With customization, users intervene and make choices.
What type of application is best suited for customization? Anything that simplifies tasks that traditionally require multiple platforms, interfaces and even user memory (ahem, sticky notes). By providing a one-stop platform for doing things your way, applications that rely on customization tend to do well for users needing active control.
Personalization: When the system makes the choices
Personalization is about the system making decisions for us. It’s far more scalable – and a very different experience. Not only does the system need to know a lot about you to begin with; it also needs to be ready to learn more as you use it. If you don’t know much about your users, opportunities to personalize are minimal.
The truth is, there are tasks for which the active control found in customized applications is not ideal. In cases in which outcome options are numerous (or endless) and user choices change frequently, automation and personalization can provide a better solution.
To illustrate, let’s return to music services. Spotify competitor Pandora Radio is powered by the Music Genome Project. While Spotify embraces customization, Pandora is all about personalization. Its machine-learning algorithm guides listeners to new music the system believes they’ll enjoy. Personalization also leads diners to great restaurants that match their tastes. Or applications — email, news aggregators, online shopping, the list goes on — that free users to filter out noise and instead shape an experience based on their own preferences.
Do those preference change over time? Or seasonally? Are users’ preferences best met by understanding their behavior and making selections passively for them? Can we learn from similar users and update UX accordingly? These are the sweet spots of personalization — but don’t limit yourself to only these examples.
The art of customization
While customization’s lower barrier to entry is appealing, it comes with its own caveat: The UX has to be so good that users want to engage with it. Overkill with customization can take users back where they started — doing too much on their own to reach their goal. If your company is building a home décor planner that relies on customization to design rooms and personalization to predict items a user may purchase, customization efforts may fail fast if they prove too complicated and time consuming. So how you present information is critical. Add too many details — or leave things too open ended — and the experience becomes confusing for customers.
Let’s talk mobile. Spotify has done a great job with customization by allowing playlist addition with a simple button in a full-screen “currently playing” view. Hit the check, add it to the list. If customization defines your product, making it easy to use across platforms is critical.
Google Gmail recognized that some administrative tasks in the mobile application provided a clunky UX. Quickly filtering and prioritizing email is central to Gmail’s differentiation, but mobile users had a hard time with this. Google’s successful solution? Let users self-define swipe actions to automate tasks they need most.
UX trends you need to know
What trends do we see emerging? The ongoing improvement of machine learning (ML) and, more broadly, cognitive computing is undoubtedly a movement worth paying attention to. Building personalized systems is easier than ever, and there’s a growing pool of talent. The advent of ML (colloquially, AI) is also challenging the view that customization is easier to adopt. Many personalized applications today are feature-light and rely solely on passive user involvement — considerably easier than creating a robust platform built on customization.
Four steps to determining the right UX
Here’s how to determine whether customization or personalization best suits your product or service goals.
1. Understand the context.
Before you embark on one path or the other, consider what you’re building, your audience and how fast you can build the product.
Consider the highly publicized new personalized menus in McDonald’s drive-through windows. Will they generate more visits in the long term? Will they increase average purchase value? Will they be perceived as too creepy? This is an important conversation — especially when your experience hasn’t previously been based on personalization, but instead customization.
Likewise, customization-driven platforms often begin building features not driven by demand or business value. New features are added and not well communicated to customers. They do little to drive sales and get lost in the pricing-page shuffle.
We often poorly understand context when the design function within organizations is isolated in the creative process and fails to take into account how or why to build.
2. Strike the right balance.
Determining the best UX option requires choosing the right mix of user interface (UI) complexity, process agility, and access to data.
- Complexity of the UI. How intuitive is the UI? What’s its maturity? UXs that are too hard result in adoption attrition. Always keep the end user in mind. The goal is to keep things simple and focus on the content hierarchy so the user isn’t overwhelmed with information.
- Process agility in engineering. At their cores, both personalization and customization are engineering problems influenced by design. To ensure a UX that continuously delivers value and business outcomes, it’s important to have a backlog of features that’s always being re-evaluated. Make sure the features you release offer the best value to end users and to the business.
- Access to user data. Share as much product data from sales and marketing as possible with the engineering team. The more data they have, the better they know and understand the end user — and the better the products and services they develop.
3. Know what your product is.
It sounds simple, but it’s common for each business function within an organization to have its own idea of what the product is. Business development’s focus is the customer-facing side; engineering is concerned with how the product is going to work. Disconnects occur when no overall plan is shared. Without an agreed upon plan, engineering may build features out of sync with the market, and business development may remain unaware of the development effort needed to bring a feature to market. A successful UX means eliminating these silos.
4. Finally, how fast can you change features?
Before making changes or updates, determine their importance to end users — or the business — and whether the time is right to release new features.
- What’s the overall effort required to make the update? Feasibility assessment requires discussions with front- and back-end engineers and UI designers.
- What are the dependencies to support the new feature? Every time you add a feature, you create another object that’s dependent on other things.
- What’s the goal for the feature? Is there customer research to support it? Prototype validation? Feedback from marketing and strategy?
- What is the viability or potential profit the feature brings?
This article was written by Andres Angelani, CEO of Cognizant Softvision. For more information, read Angelani’s new book, Transforming While Performing: A Practical Guide to Being Digital. Learn more about software product engineering by visiting our website.