Business Challenge

Acquiring new customers is one of the most challenging investments for a scaling business. The relationship is unproven: customers have yet to demonstrate their long-term commercial value, while the business has yet to earn their trust and loyalty.

The challenge is not simply converting a first purchase, but identifying which relationships have the potential to grow—and creating the conditions for customers to return.

At Farfetch, elevated VIP service was traditionally reserved for customers who had already demonstrated their value through spending. This created a missed opportunity to build relationships with high-potential customers earlier in their lifecycle.

We explored a different approach: using predictive intelligence to identify customers with the greatest potential for long-term value, then investing in personalized human service before that value had been realized.

The business hypothesis was that earlier, targeted engagement could accelerate repurchasing, increase loyalty progression, and reduce costly friction—creating greater value for both the customer and the business.

Designing the Personal Service Advisor

Working with my research and design team in partnership with Customer Service, I led the development of the Personal Service Advisor concept: a proactive, concierge-style offering for newly identified high-potential customers.

A new predictive algorithm provided a means of identifying customers with strong potential to progress into Farfetch's highest loyalty tiers.

The service concept translated that prediction into a tangible customer experience. Advisors could establish a personal relationship early, provide tailored guidance, anticipate needs, and intervene where additional support could influence purchasing confidence and loyalty.

The proposition combined several forms of high-value support, including personalized product recommendations, size and fit guidance, assistance with first purchases, and support through cancellations or other moments of uncertainty.

The objective was not simply to improve satisfaction with Customer Service. It was to test whether the combination of predictive intelligence and human-led intervention could influence purchasing behavior, accelerate loyalty progression, and reduce avoidable service costs.

Designing and Testing the Experience

We began with a low-fidelity proof of concept, working with three Customer Service advisors across the United States and Mexico.

The six-month pilot was designed to evaluate the commercial and operational viability of the model against a control group, focusing on three questions:

• Could participating customers progress into higher loyalty tiers more quickly?

• Would personalized intervention increase purchasing frequency and customer spend?

• Could targeted assistance reduce costly friction, including returns?

An equally important question emerged through testing: how could proactive service be delivered in a way that respected customer preferences while remaining economically scalable?

Early outreach exposed limitations in the initial channel strategy.

Phone contact offered the potential for high-touch engagement, but proved ineffective for establishing relationships with customers who had not requested assistance. Only five of the first 300 customers responded to invitations delivered by phone.

Synchronous chat presented similar challenges, requiring customers and advisors to be available at the same moment.

We adapted the approach, shifting to personalized email outreach that allowed customers to engage on their own terms while preserving the individual advisor relationship.

Participation increased tenfold following the change in outreach channel.

This was a critical learning: the value of personalized service depended not only on the assistance offered, but on how and when customers were invited to engage.

Throughout the pilot, customer feedback, advisor observations, and performance measures informed iterative improvements to the service experience.