Adoption Is an Emotion, Not a Metric
The seed for this all started back in February when I hosted Adam Treitler on the monthly HR Tech Couch Club where we take an HR practitioner’s lens to all things HR tech (you can watch the full episode here). In that session, Adam said something about how AI will drive an increased focus on UI (User Interface) and UX (User Experience) in tech design. When I first heard it, I thought, “Hmm, interesting…” and left it at that. But two months later, I found myself still thinking about it and decided to figure out why I’m still chewing on that particular piece of mental bubble gum.
Sparing you all the mundane details of the research behind this, here is where I landed: at the core of most AI business cases is an assumption that the tool will make things more efficient. It is that primary goal for efficiency that is preventing the systematic and sustained adoption of AI across the enterprise. Executives may believe in the value of AI, but a belief in its business value doesn't mean automatic adoption across the enterprise, and adoption doesn’t usually happen until people want to use the tool.
This is where I believe we are taking another Internet Era approach to an AI Era problem. Prior technology shifts used to be about function (i.e., What can the technology do? How much money can it save? Etc.), whereas AI is fundamentally about communication. It’s the communication between humans and machines that yields a higher caliber output. This means that the design and adoption challenge here isn’t just functional (do X and yield Y); it’s conversational, emotional, and, in a way, playful.
Borrowing from what Chris Dixon wrote back in 2010, “The next big thing will start out looking like a toy,” here are some ideas and reframing that could help you when you select, design, implement, or try to drive adoption for your next enterprise AI technology:
Tool vs Toy
Quick thought experiment here: picture a kitchen spatula. How did that make you feel? Utilitarian? Are you set up to complete a task of sorts? Now, imagine a pile of Lego blocks. Did that shift your sentiment? Do you feel a sense of exploration? Maybe there is an opening of possibilities?
At their core, tools are designed to perform a task, while toys are designed to invite interaction. Whether we like it or not, they generate different sentiments across the user population. Because of this, the best “toys” in tech can disguise their complexity through delightful, engaging, and playful experiences that make users want to explore rather than comply. Take Duolingo, for example; it’s a language learning app that is wrapped with gamified streaks, bright visuals, playful sounds, and immediate feedback that results in users returning daily, not because they have to, but because there is a level of emotional satisfaction that is associated with using the app.
When we examine some behavioral design principles more closely, a toy-like experience actually makes sense. Take the Fogg Behavior Model, for example. In this model, Behavior = Motivation x Ability x Prompt. Toy-like experiences can reduce the friction in Ability and increase Motivation through design.
Ok, so what on earth does this all actually mean for you if you’re neck-deep in an AI HR tech RFP right now? Well, a few things:
Avoid tools with bland interfaces, steep learning curves, and that offer a low emotional return on interaction. Yes, it is still technology for work, but when did we become okay with the idea that work tech shouldn’t be engaging?
Play with the tool for a while and see if interactions with AI feel natural and organic. And most importantly, does it create a desire for you to continue interacting with it? People aren’t usually opposed to talking to a robot; it’s when the responses are emotionless and robotic that we see resistance to adoption
Using the tool should feel like a creative collaboration and not a data entry exercise. This is where pilots become super important. Make sure your intended end-user base get a chance to try the tool and provide some feedback to you before you move ahead with mass adoption
Now, being realistic here, I think most enterprise AI tools today are still designed to function in “tool mode” (read: bland UI, super techy, and provides zero incentive on repeat usage). This doesn’t mean that you shouldn’t implement it (after all, when was the last time HR actually had veto power on technology?), it just means that you need to account for the steeper change management and adoption curve and be realistic about your ROI timeframe.
Behavior Design
I can’t remember when it became the norm to “train” users on technology. In my career, I have seen people being trained to use HRIS, ATS, People Analytics systems, etc., and now I’m seeing it become a norm to train people on AI.
We often over-engineer tools and the adoption process for logic and under-design it for instinct. We assume that people will follow instructions and SOPs, attend trainings, and figure out how things work as long as the business value is clear. But humans don’t work like that. To drive adoption, we need to design for behavior and not just function, especially when you are trying to change how people work.
Here are some ideas we can borrow from behavioral economics when designing change programs:
Defaults Matter: People rarely change default settings. Make sure you nudge behavior by pre-selecting some helpful options (e.g., offer prompt starters, a set of curated use cases, lead with “next best actions,” etc.) to help users start without fear
Loss Aversion: People hate losing more than they love winning. Try framing inactivity as a lost opportunity (e.g., “you’re leaving 30 minutes of your precious time on the table”)
Chunking and Micro-Progress: People feel overwhelmed by large tasks, but energized by smaller ones. So instead of setting the goal of “becoming AI proficient in two months,” tackle it one small step at a time with “using AI to answer one work question today.”
Social Proof: Humans copy others. If others are doing it, we feel safer doing it too. This is why lunch-and-learns work, because you are seeing your colleagues in action. In change programs, you can also share insights like “teams like yours are using this prompt to reduce meeting time by 20%” to help drive adoption
At the end of the day, people don’t fall in love with tools just because it makes them faster, they fall in love with tools that make them feel more capable, more creative and more in control.
Empowerment in design and adoption means giving people the confidence, tools, and space to go beyond what they thought they could do and create, express, and problem-solve in their own way.
Interface is Culture
AI adoption at scale doesn’t come from features and functionalities; it comes from feelings. Specifically, feelings of trust, delight, ease, and agency. User experience is at the core of this as it shapes how people work, how they think, and how they relate to each other at work.
The interfaces you select, design, and implement become a part of a company’s shared behaviors, norms, and expectations. Over time, these behaviors shape culture.
If a tool interface is cold, rigid, or overly complex, it communicates to the users that their work needs to follow those same patterns; whereas if a tool interface is open, playful, and adaptive, it gives permission for users to experiment, be curious, and explore possibilities.
Research done by the MIT Media Lab concluded that emotion is not peripheral to technology design but rather central to how we experience and adopt technology. UX that acknowledges human emotional states (e.g., frustration, delight, curiosity, etc.) resulted in more engagement and loyalty.
In the book, Emotional Design, Don Norman showcases how visceral, behavioral, and reflective design elements shape emotion and memory. In other words, attractive things work better because people approach them more openly and patiently. Users tend to form long-term attachments to products that make them feel good, not just perform well.
What does this mean for you? Understanding that emotions drive repeat usage, form team norms, and ultimately drive organizational culture will help you shape how you might approach technology customization and give you additional angles to drive change management and adoption of AI tools across the enterprise. For example:
Instead of talking about what tasks the tool can do, talk about how the users will feel while using the tool to complete the task
Instead of focusing the narrative on speed, business value and logic, focus on creating confidence, curiosity and clarity
Instead of assuming end-user expertise in your change management approach, assume end-user fear and hesitation
Create an environment where the user collaborates with the tool and the rollout process instead of just complying with it
A functional user experience gets you through the task, but an emotional experience makes you want to come back. When you start with feelings and not features, you will start to unlock the magic of sustained AI adoption across the enterprise.
I know it may feel a bit counterintuitive to talk about feelings and emotions (aka the “mushy” things) in the same conversation as technology design and adoption, but I genuinely believe that the next wave of breakthroughs in tech is focused on designing for human sentiments.