Ethical design: growth through fairness

Ethical design is the practice of building products and journeys that respect four things customers cannot get back: their privacy, their choice, their time, and their effort. The everyday counterexample is the cookie banner. Accepting is instant; refusing takes a hunt through settings. The interface is engineered to induce fatigue, and it works hardest on the most vulnerable users. Consent without a real choice is not consent. The same test applies to any tracking justified as helping customers: did they ask for the help, do they have genuine freedom to say no, and does the context they shared data in match the context you are using it in?

Why it matters to the business

Personalization done with permission pays. McKinsey's Next in Personalization research found 71% of consumers expect personalized interactions, 76% get frustrated without them, and personalization most often drives a 10 to 15% revenue lift. But the lift compounds only when customers accept the exchange.

Done without permission, it backfires. Accenture found that among consumers who felt an experience was too invasive, 64% blamed brands using data they had not knowingly shared, and its Global Consumer Pulse found 41% of US consumers switched companies over poor personalization and lack of trust, a cost it put near $756 billion a year for US organizations. Cisco's 2024 privacy survey adds that 75% of consumers will not buy from organizations they do not trust with their data.

How to use it

• Build personalization on first-party data customers knowingly shared. Retire data they would be surprised you hold.
• Explain why each recommendation appears, in one plain sentence the customer can see.
• Ship a preference center with a real, easy opt-out, and make 'no' as fast as 'yes' on every consent screen.
• Test consent flows with your least digitally confident customers, not your power users.
• Before any new tracking, answer three questions in writing: did they ask, can they refuse, does the context fit?

Where teams get it wrong

Teams treat 'customers want relevance' as blanket cover for surveillance. Optimization is not understanding: an algorithm knowing what someone clicked is not the same as caring about their problem, and algorithms carry no context, empathy, or responsibility for the outcome. Creepy relevance destroys the trust that makes relevance profitable, and a design guideline in one team cannot fix that. It takes governance, accountability, and transparency across the whole organization.

Ask your team

• Which of our personalization features use data a customer would be surprised we have?
• On our consent screens, how many taps does it take to say no versus yes?
• Could we explain every recommendation we make to the customer's face?

Optimization is not understanding.