Product Thinking
Personalization Starts With Journey Awareness
Relevant personalization begins by understanding a person’s current intent and stage, not by adding more content or recommendations everywhere.
· Jason Wong · 2 min read
- Digital Products and Service Design
- Practical framework
Personalization often begins with a catalogue: which offer, article, destination, or product should we show? The better starting point is the customer’s moment. Are they exploring, comparing, planning, transacting, preparing, or recovering from an issue?
Relevance depends on intent
Journey awareness gives personalization a reason to exist. It helps teams decide when to be helpful, when to stay quiet, and what information someone needs next. A customer who is exploring may need inspiration and confidence. Someone comparing options may need clarity about trade-offs. A person who has already committed may need preparation, reassurance, or service support. Showing the same content to all three is not personalization; it is distribution. The work begins by defining the journey stages that matter, identifying the signals that make each stage credible, and connecting each stage to a real customer need. It should also include a fallback for uncertainty. Signals are imperfect, and an experience should never become intrusive because the system has guessed wrong. More content does not create relevance. A better understanding of the moment does.
Expectations are high, but the signal is imperfect
McKinsey’s 2021 survey found that 71 percent of consumers expected personalized interactions and 76 percent were frustrated when they did not receive them. Yet the same commercial pressure can encourage teams to act on weak signals or over-personalize. (McKinsey, 2021) Journey awareness provides a control. It forces every use case to connect the signal to an inferred need and a suitable action. A useful decision table includes:
- Stage: What is the person likely trying to accomplish?
- Evidence: Which behaviours make that inference credible?
- Confidence: How certain is the classification?
- Response: What help is appropriate at this stage?
- Fallback: What neutral experience appears when confidence is low?
- Measure: Did the intervention improve the journey without increasing avoidance or opt-out? This model prevents personalization from becoming a contest to show that the organization knows something. The objective is not recognition for its own sake. It is useful help at the right moment.