Maintaining user confidence in evolving tech products
For companies offering consumer products, innovation is key to unlocking new value for consumers and staying relevant among the competition. But when does innovation turn from impressive to intimidating? Particularly for habitual product users, sudden and large changes can be overwhelming. Users may fail to recognize the additional value of the new features or find that they disrupt their existing habits. How can companies determine and sustain an innovation velocity that consistently takes advantage of new technological advancements to maximize potential product value while ensuring that users maintain a strong understanding and comfort with the product?
This summer, I worked at a BMW dealership as a Genius (aka Product Specialist) where I had the opportunity to learn both perspectives. In my technical training, I learned about the history of innovation in the automotive industry. In my customer interactions, I learned how these innovations have affected customers and their usability — both for better and worse.
My role uniquely placed me in a position to observe hundreds of human-computer interactions. Below I detail encounters that most surprised me, including features that flopped in the automotive industry, and my recommendations for consumer tech products to more effectively balance innovation and user confidence moving forward.
Disclosure: Despite primarily selling BMW vehicles, the dealership where I worked sold many car brands. My interactions with customers reflect the issues I witnessed across the entire automotive industry in many different makes and models, not specifically BMWs.
Fear and uncertainty are powerful deterrents
In my first-ever customer interaction, I had a customer tell me she was ‘too scared’ to use her car. She leaned in and whispered to me, “I think something may be wrong with it.” As a result of her fear, she refused to use the car for over a week! That seemed like a major cause for concern, and I told her I’d look into it right away.
What was the issue? Her primary concern was that the steering wheel was shaking. As the previous owner of a problematic ’98 Saab with an alignment issue resulting in the same symptom, I was empathetic to her concern. So I took the car for a drive to confirm the issue.
As it turns out, the issue wasn’t mechanical. And it wasn’t even an issue. The shaking was simply haptic lane assist feedback: if you drive slightly out of the lane lines, the steering wheel vibrates to prompt you to make an adjustment. That’s it. It amazed and worried me that a new feature meant to help the user and inspire safety confidence was causing so much distress that it resulted in a refusal to use the product entirely.
Not only was this customer failing to recognize the value of a new(er) feature, but its presence distracted from her expected experience in a way that sparked confusion and fear. Distractions aren’t inherently bad; many find this interference to their driving (through lane assist feedback) helpful, given that it trains them to drive more safely. However, the positive response relies on the fact that users can immediately understand the meaning of the feedback and how they should react. The customer in the above scenario never made this connection, resulting in a negative response and, ultimately, perceiving a larger issue that didn’t exist.
When customers can’t explain why something is happening, they’ll postulate reasons of their own. In the above scenario, a single user assumption was enough to deter product use. And although I was able to provide clarification and resolve the perceived issue, a customer might feel that it isn’t worth the effort to reach out and instead continue with their false beliefs. In this case, the customer may have never touched her car again!
The Genius role of providing customer support is built into the automotive business model to provide these clarifications to customers. But to what extent should it need to be? How much teaching and clarifying should be necessary, especially for an everyday product? Shouldn’t users be able to discover and teach themselves these new features?
Self-discovery can be dangerous
Over the course of the summer, I witnessed a number of other misunderstandings and frustrations, as well as excitements and delights. I tried to ‘facilitate’ user learning, stepping in where I could be helpful while allowing the customer to take the reins and gain autonomy over their vehicle. Here’s how it went…
👎 I heard about a driver who rolled into a pole on their first day of owning their new car. Clumsy or just confused? I learned that the driver was unfamiliar with the design of the gear shifter and consequently made assumptions about how it worked. Well, at least one of those assumptions was false (i.e. how to put the car in park), and this was the result.
As a Genius, I ensured that every customer I worked with knew how the gear shifter worked before driving (specifically to avoid this). Many were, in fact, surprised to learn how it worked compared to their previous cars. It’s pretty intimidating when the most fundamental operations change, and it can result in serious consequences.
Several types of vehicle gear selector/shifter interfaces, each with their own interaction paradigms
👍 I watched users’ delight in discovering CarPlay for the first time. This was a frequent experience, and you would think I was truly a genius based on their reactions. Some customers had the feature available to them but had never tried it until I showed them how. Most users recognized the value as soon as they experienced the feature, however, many didn’t know about it or think it was worth the initial effort to set it up. How might the car lower the barrier to trying these features or help users more quickly recognize their value?
👎 … and the frustration if CarPlay stopped working. For every customer who discovered their newfound love for the feature, there was another who had already learned to love it and came in fuming that it wasn’t working as they expected. The customers who gained the feature and, as a result of some technical glitch, had it taken away were significantly more dissatisfied than those who never knew about it in the first place. It’s important that once innovations change a user’s daily routine, they must consistently work as expected. Once a routine is built, variations in the process can be very frustrating.
👍 I watched users marvel at features like gesture control and touchpad writing. Many users don’t know about these features, but they’re pretty cool. A small selling point for those looking for the ‘luxury’ experience. As a Genius, they were some of my go-to features I liked to show off how ‘cool’ the car was.
👎 … and users who unexpectedly discovered these features on their own. One time, much to the driver’s dismay, the car suddenly started blasting music because the driver had accidentally triggered radio gesture control by talking with their hands.
Several times, drivers were trying to input their address into the nav system but instead were getting characters like “,.-~” because the touchpad was trying to recognize characters from their tap, rather than recognizing they wanted to push to ‘select’ the character on-screen. In both cases, these features were getting in the way of what the driver was trying to do; they created more confusion than value when they popped up unexpectedly.
Across the board, users often failed to self-discover features or did so unexpectedly to their dismay. How might companies enable user discovery in a productive way to help them recognize the value of new features?
User trust and comfort will be paramount with the introduction of AI solutions
Many of my observations addressed customer learning of product innovations that have been introduced over the past ten years. But what about the next ten years?
Innovation in the automotive industry, and more generally in the consumer tech product space, has the potential to grow exponentially in the next decade with technology advancements in data processing and artificial intelligence (AI). And this may create some major advancements in product value.
“Autonomous driving (AD) could revolutionize the way consumers experience mobility. AD systems may make driving safer, more convenient, and more enjoyable.” — McKinsey & Company
But if users are scared of lane assist feedback, how do you think they’ll handle autonomous driving features? While many think of full automation when considering AD, there are various ‘levels’ of AD. Cars with lane assistance are level 1 (and sometimes 2), and automotive innovation will soon progress through conditional automation (level 3) before reaching full automation (levels 4 & 5).
Levels of Driving Automation, summarized by automation company Synopsys
I predict that level 3 will be a scary leap for many drivers. It switches from human monitoring to automated system monitoring while still requiring driver attention and manual takeover at a moment’s notice. How can a driver be sure when they are controlling the car versus the automated system? How will the system warn and prompt them to take action in a way that doesn’t overwhelm them?
With the ongoing innovations of autonomous driving in the automotive industry and AI features across other industries, it will be imperative that systems are designed to enable user understanding and comfort to create long-term trust. If companies fail to do so, they will instead see resistance to these features (and the prospect of future innovations) that are supposed to be driving value.
Recommended design strategies for consumer tech product innovation
Companies shouldn’t halt their product innovation; they just need to introduce new features in a way that will build and maintain trust. Here’s what they need to keep in mind:
1. Don’t interfere with the core processes, or make it abundantly clear if you do.
In a car, for instance, people need to be able to change gears, adjust their seats and mirrors, put on their seat belts, steer, and park. Companies should avoid introducing new features or changing existing features that may interfere with the existing routines associated with these core processes.
Especially as ‘cool’ or ‘luxury’ features are added, the additional value will only be recognized if they are in addition to the existing features driving the primary value. If some of these systems are used to replace the existing features, this change needs to be transparent and thoroughly explained to update the users’ expectations.
In the replacement scenario, the design also needs to protect against the potential hazards of a user performing their old routine. When I was learning to drive, my instructor warned the class about a girl who accidentally reversed on the highway while trying to turn on her wipers. She had just bought a new car where the gear shifter was a lever in the same position as the wiper lever in her old car. My instructor used this to emphasize the importance of familiarizing yourself with your vehicle, but it also exemplifies a poor design choice that was not protecting against the hazards of an old routine.
But really, why replace for the sake of replacing? Don’t reinvent the wheel. Or the gear shifter. Unless there’s really a need to. If a current core process solution is working well for the user, companies should be wary of changing it.
2. Ensure that feedback can quickly be attributed to the action it’s reinforcing.
As I’ve discussed, people are generally averse to interference with their routines, but it can be helpful to provide user feedback in certain scenarios. Lane assistance, for instance, is valuable feedback if perceived correctly. The issue in the initial scenario was that the car provided no indication that the vibrating steering wheel was attributed to lane assistance, so the driver couldn’t connect the dots.
Feedback is only effective if it’s clear what action it is addressing. To do this, feedback must be delivered 1) immediately, and 2) with proper contextual information. If either is missing, the additional light/noise/movement/etc. is truly a distracting interference. It adds to a user’s confusion and fails to change their future behavior.
3. Use guidance purposefully to help users self-discover features at the right times.
From my customer observations, I found that users often failed to self-discover features, or did so unexpectedly to their dismay.
If you want users to recognize the value of the features you’re offering, you need to help them experience that value. It fits into the product-led growth movement among many software companies — driving value from the product itself rather than marketing or sales efforts.
But sometimes, features may hold different value at different times. Massaging seats, for instance, can be jarring if accidentally activated in a high-traffic area where the driver needs to be extra attentive to their surroundings. When you’re sitting in an idle car, meanwhile, they can be great!
The product design should prompt feature discovery at times when it would enhance the current experience and, conversely, should mitigate feature discovery when it would detract from the experience. Many drivers never try cruise control because they don’t know when or how to activate it. Perhaps the product could temporarily ‘highlight’ relevant features and ‘hide’ non-relevant features to guide the user’s actions. Or offer a tutorial at a time that would be convenient. Give the user what they would want when they would want it.
4. Perform usability testing across demographics.
To discover how users really perceive potential product innovations, companies need to put the new features in front of customers and perform usability testing. This creates opportunities to discover issues (like those I observed in my role) before they become issues. As Robert A. Virzi reported (and later echoed by Jakob Nielsen and Thomas K. Landauer), “80% of the usability problems are detected with four or five subjects.” Discovering these issues and misperceptions early on is critical.
However, you won’t encounter 80% of usability problems unless you accurately sample your user base. One of the simplest considerations is sampling from different age ranges. Do usability assumptions hold for both younger and older user demographics?
You might alternatively consider a demographic sampling based on the version or brand of product a user is upgrading from. Each time I helped a customer transition from their old vehicle to a new vehicle, I made sure to familiarize myself with their old vehicle so that I could point out specifically what interactions had changed.
A customer switching from the second-latest model to the latest model likely won’t experience much trouble, but how about a customer who hasn’t upgraded in 10 years? Or a customer coming from a competitor’s product? Testing across demographics ensures that you maximize usability problem discovery, even if only testing with a few users.
Conclusion: bring your customers along with you
While my observations from the automotive industry made me recognize the issue of imbalance between innovation and user understanding, the resulting recommendations generalize beyond the automotive industry.
For any consumer tech product, companies must be careful to maintain user understanding as they innovate. In some cases, they already need to regain lost understanding. False user assumptions hinder usability and lead users to form negative perceptions of the product over time. And users only make more false assumptions as they lose understanding of their control over the product.
While many companies are eager to integrate technological advancements in ‘exciting’ new product features, it’s important to remember that ‘excitement’ isn’t inherent. I hope that by considering the proposed design strategies, companies can bring their customers along with their vision more effectively, reducing friction and helping them experience this excitement, too.
This article was modified from its original October 1, 2024 publication in the ‘Designing for Empathy’ issue of the ‘Interaction Nerds by UXA’ newsletter. I’m one of several Carnegie Mellon undergraduate students writing for the weekly newsletter — if you like this article, you should check out the rest of our work!
