Trust in new technologies: 5 guiding principles
Everyone probably knows the hype around self-driving cars. A cool technology that can totally change the way we commute. Or machine learning that can predict our behaviour, make decisions on behalf of us or automate certain tasks.
How far can we go? Everything is possible nowadays. The sky is the limit. But good products tend to have more qualities than just the technology itself. By listening closely to our users, we create solutions that people are actually willing to use. Basing your product on a certain (latent) need leads to success. Always. But, is this applicable in every situation? If you would ask me a year ago, I would say yes, definitely. But experience thought me that there is more to it, especially regarding new technologies. Namely, trust.
Why? Well, trust equals adoption. As a user, gaining trust means starting to adopt the new feature, piece of technology or functionality. This is done by trying it out or daring to do, also known as ‘taking a leap of faith’. If enough trust is gained, the user feels comfortable to start using it. On the other side of the coin, it is extremely important to not lose trust. As written in literature, bad experiences lead to a bigger decrease of trust than an increase of trust in good experiences. To make it concrete, would you step into a self-driving car any time soon if you just saw one crashing? I don’t think so. That’s why we have to guide our users to trusting our solutions. Once people start distrusting our product, we are a long way from home.
In the end, it is up to the user to adopt the newly released functionalities and thus unfamiliar technologies. Ensuring adoption and success can be done by designing for the appropriate trust and experience. The following five principles can guide you to do just that.
Eliminate bad experiences
It may sound obvious to never intend for bad experience, but trust can easily be damaged by experiences that were not as expected. For example, mistakes that were made by the system accidentally. Think about wrongly predicted elements in your software like a weather or traffic forecast or a route planner sending you to the wrong place. When such thing occurs, people tend to greatly distrust the system and it becomes harder to adjust the trust to normal levels again.
How to solve this? Well, start with spotting a bad experience when it happens. Think about what could go wrong and start building monitoring tools in your software. Keeping track of how your users use the functionality or how they navigate through your site can tell you a lot about unintended bad moments. To be even more specific, in a project I personally was involved in, we monitored whether our users followed up on our suggestion. And if automatically filled in information was changed later on. This didn’t only tell us how good our predictions were, it taught us also how trustworthy we are.
Use small steps
To increase or gain someone’s trust, this has to go step by step. As mentioned, people tend to take a “leap of faith” and when this was a successful experience the trust is a little bit increased. Let’s take this to a day-to-day situation. Imagine yourself eating something weird for the first time, you have to trust that it will taste good. You dare to take a small bite, and … Was it nice? You want to take another bite? Similar to experiencing new food, these are the small steps we have to incorporate in our solutions to gain trust and ensure adoption.
To give an example on how to incorporate small steps related to the industry I’m personally involved in: automated bookkeeping. Doing the books used to be a quite labour-intensive task. Luckily, we are now able to automate more and more. Based on historical data, we can for example predict what G/L account should be booked on. But before we can go to full automation, and allocating G/L accounts without the users’ interference, we have to guide them in order to fully trust the system to do it. Therefore, suggestions were created so the user can decide to follow up in our advice or not. Slowly, we can move to a situation where our advice can be trusted and automatically filled in.
Give control
What do you do when you don’t trust a situation? Or you don’t know what is going to happen if you press this button? You seek control. You want to be in charge. It can be expressed via control settings, think about choosing a program on your washing machine. You know what to expect when you wash your fragile clothes on 90 degrees Celsius. Or, being able to choose. For example, when suggestions are presented, you can actively decide to use or ignore that information or choice.
In what other ways can you give control? It’s also about knowing what’s going on. It is about ‘being in control’. To give an example that entrepreneurs might relate to, is (again) bookkeeping or financial administration. A lot is happening in the background. New bank statements come in, and sales are made. Providing trust and control in such situation is guided with dashboards and alerts. With such tools, someone can for example ‘be in control’ of their expenses. But also showing aggregated numbers and highlight outliers in other situations where a lot is going in the background can be perceived as trustworthy to the end-user.
Provide confirmation
When you first use a new thing and thus taking this ‘leap of faith’, you want monitor what just happened. What was the result of your action? Did it go well? Providing confirmation and feedback can help in sustaining the trust and create confidence. Think about in screen feedback, showing the result of clicking a button, or changing a status from “draft” to “send”.
To be more concrete, it is about confirmation to avoid hesitation from the end user. What is it that your users are not sure of? Especially the first time they use it? To give an example, electronic invoicing is a new method to invoice your customer as an entrepreneur. What you want to be sure of in this process is that the invoice will actually be delivered and ends up at the right person. The main reason for this, is to get paid in time. Knowing this, you can think of solutions that provide confidence. Like for example knowing that the receiver of the invoice wants to collect invoices electronically. Or presenting a detailed delivery status just after the invoice was sent out. There is no reason to worry if you know from the status that the invoice is already picked up and processed. Likely that your payment is already on its way.
Think about potential risks
People are cautious to dare to take this ‘leap of faith’ when it concerns something with high risk. Think about driving a car on high speed, sending e-mails on behalf of you, or mobile banking for elderly people afraid to lose their saved money. How to deal with this? Focus on the less risky elements, provide the right format that is perceived as trustworthy (e.g. human interaction for elderly people) or highlight potential risk factors.
So, what is risky for your customers? What would be hurting them if it goes wrong? What is hard to change afterwards? An example could be filing your tax return or sending a payment. Once processed means no way to get it back. Therefore, messages like “are you sure?” where you’re able to review and confirm are good ways to increase trust in the system.
