Design Design Design! → Part 95
What is user personalisation — and how can AI help?
What is personalisation (really)?
Personalisation in UX means shaping content, features, or experiences around a specific user’s needs, preferences, and behavior — often based on data. Done right, it makes products feel smarter, faster, and more relevant. Done wrong, it can feel annoying, creepy, or just plain boring.
Why it matters
- 77% of users prefer brands that personalize their experience. (source)
- People now expect interfaces to “get them.”
- It’s a competitive edge — but also a risky game if overdone.
Personalization vs. Customization
Quick distinction:
- Personalization = system-driven (based on data)
- Customization = user-driven (based on choices)
Both are useful. Both can go sideways.
Where personalization shines (2025 examples)
- Streaming: Netflix recommends what you’re likely to watch next.
- Social Media: Meta filters your feed based on recent behavior.
- E-commerce: Amazon tailors offers using user data + ML.
The dark side of overpersonalization
1. Lack of diversity
Too much focus = tunnel vision. Users get stuck in content bubbles.
2. Homogeneous experience
If everyone gets the same kind of “personal” content, it all feels the same.
3. Redundancy
Same ads. Same products. Same everything. Users bounce.
4. The creepy factor
Personalized content based on unknown data trails = distrust.
5. Overcustomization → cognitive overload
Letting users tweak everything = too much work. Keep it simple.
How to personalize (without messing it up)
✅ Balance relevance with variety
Show what users want — but also what they might not know they need.
✅ Blend general + targeted content
Give room for discovery, not just prediction.
✅ Respect sensitive data
Context is everything. Don’t surprise users with personal info.
✅ Give control
Let users manage what’s being personalized (and how).
✅ Listen to feedback
Let users flag bad recommendations or creepy content.
✅ Limit customization
Offer choice, not chaos. Focus on impactful custom features only.
When it works: A quick example
With Acadeum, we built a brief onboarding flow that asks users what they’re looking for. That data fuels smart course recommendations later on. No guesswork. No overkill. Just helpful UX.
How AI helps with personalization
AI supercharges personalization by identifying patterns no human could catch. It processes behavior, predicts needs, and adapts content in real time. Think: smarter recommendations, better timing, and continuous learning. But it should always serve the user — not just the algorithm.
