How do you design AI products people can trust?
You probably know the answer already. But stick with this.
Every AI product leader dreams of shipping an AI feature that customers actually want. The problem is most teams don't know how to connect with customers. They just assume they know what people want.
Hint: They don't.
87% want AI disclosure when interacting with technology. That is not a suggestion, that's a mandate.
The only way to know for sure is to ask your customers. Which means talking to real humans and gathering insights about their experiences.
Time to break this down…
Table Of Contents:
Why Ethical AI Starts With The Customer
Launching AI features without talking to customers is like walking blindfolded. You might make it, but you'll probably stumble along the way.
Trust in AI technology is already shaky these days.
82% fear AI data loss as a personal threat. A separate study conducted by Prosper Insights found that privacy violation due to AI concerns nearly 56% of US adults who are either extremely concerned or very concerned.
Now, imagine what happens when your shiny new feature violates that trust. You've instantly destroyed months (or years) of goodwill by not talking to customers.
Your competitors are launching AI features.
Regulators are starting to take notice.
Users are freaking out.
If you build features without real customer insights your products will..
…and that's just the beginning.
Feature teams need to know what customers want…and why. Only user feedback will tell you that.
What Customer Feedback Platforms Actually Do
Customer feedback platforms are solutions that allow you to recruit, interview and collect insights from your users.
They come in all shapes and sizes. Some focus on lightweight surveys, others are deeply focused on user research. UserInterviews, for example, is built around helping teams recruit users for their research studies.
Good customer feedback platforms will allow you to..
Here's the kicker… data on its own isn't valuable. Only actionable insight is useful.
When selecting a customer feedback platform keep your goals in mind. As an AI team, you'll want a tool that helps you collect feedback ethically (and quickly) without sacrificing user privacy.
Think about it… If you skimp on your research platform, you'll end up with subpar data. Garbage in, garbage out.
How To Collect Insights Without Breaking Trust
This is where most teams go wrong.
Companies collect customer data willy nilly. Then act surprised when users unsubscribe.
Avoid this by establishing some ground rules for gathering ethical customer feedback.
Be Upfront About AI Use
Whenever possible let users know your product is using AI. It doesn't need to be a big splashy announcement, just a quick mention works.
Deloitte's 20th Annual Connected Consumer Survey found users already believe technology is improving faster than necessary safeguards can be developed. Make your customers feel safe by being transparent about how AI impacts their experiences.
Users will respect you being honest with them. Which leads to better feedback.
Get Real Consent
Before you collect data from users you should obtain proper consent. Which means..
Real consent:
If your legal team can't explain your consent form in one sentence; they need to rewrite it.
Pay Participants
Do you ask your users for their time, you should be paying them for their time. Fair pay:
When you pay for participants you start to see a wide variety of viewpoints. When you rely on "free" feedback. You usually just end up hearing from people who want to complain.
Protect User Data
Encryption doesn't happen just because you say it's important. Take steps to protect your customer data.
That means:
Users can forgive iOS storing location data but they will never forgive you if you lose their private data.
Turning Feedback Into Ethical AI Features
Now that you've got this process down it's time to start implementing feedback.
Build Features From User Pain Points
UX strategy often starts with problem statement. This is your AI feature bible.
Use insights gathered from your platform to determine what users are complaining about the most. Then rank those problems from most painful to least painful.
Solve one of those problems with your next AI feature.
Prototype Small
Don't launch your entire feature. Create a small experience that leverages AI. Then gauge customer sentiment.
Questions to ask:
If you answered no to any of these questions, solve that pain point before you go big.
Bake Ethics In From The Start
Ethics isn't a checkbox on your roadmap.
It's a culture you instill from the start.
Conducting ethical AI research means…
Pushing these steps to finish rarely ends well. Eventually, your team will be forced to go back and "ethically" collect all that data you already stole.
Remember: Your feature will never be done.
Keep The Feedback Channels Open
Every AI feature you build is never truly finished.
AI products learn and evolve over time. Which means you need to make sure your team can accept feedback about AI glitches or mistakes.
The best customer feedback tools make it easy to allow users to report pain points right inside your product. That means you can push AI feature updates without needing to kick off a new round of interviews.
Teams that listen to customers while AI is running end up with better features over time. Whereas teams who ignore customer feedback cause their features to slowly get worse (and unsubscribe users).
Bring It All Together
Building ethical AI features isn't hard. But it does take discipline.
You need to talk to your customers. Listen to what they say. Then act on it.
Customer feedback platforms give you the means to do this at scale. With the right platform you can…
Users are skeptical of AI. They should be.
With so much press about AI bias and unethical uses of AI. Your team needs to make extra effort to gain their trust.
If your team is willing to talk to customers and act on their insights. You'll have a product that not only earns trust… but maintains it.
Start with a single user interview. See how customers feel about your current AI features. Then act on what you learn.
That's how you win trust, one interview at a time.
