The best service is the one the customer didn't have to ask for
Channels working independently, with no integrated view of post-purchase.
I facilitated research and mapping with every side, from stores to the call center.
Physical stores and online shared a common pain point: delivery.
A proactive service platform, still in use today.
The challenge
The company's channels worked independently. Physical stores, e-commerce and customer service each had their own processes, and no one saw the post-purchase experience as a whole. This created friction for customers and rework for support.
We started from a hypothesis: if we acted before customers reached out to support, we could reduce friction, increase satisfaction and also bring efficiency and cost savings to the business.
Exploration
The method changed depending on the question we needed to answer. I visited stores in different neighborhoods, because the audience varied a lot from one place to another and I wanted to find what they all had in common, without generalizing from a single store.
In parallel, we listened to call center calls to understand the main reasons for contact and patterns of dissatisfaction, with both qualitative and quantitative analysis. We also analyzed call center service data. With each channel's team, I mapped the service flows to identify steps that could be integrated or standardized.
- Visits to stores with different audience profiles
- Call center call listening, qualitative and quantitative
- Mapping each channel's flows with its own teams
- Personas and journeys across every side: consumers, stores, call center tiers and stakeholders
Synthesis
By crossing the visits, the calls and the flows, we identified which situations could be anticipated instead of waiting for the customer to reach out. The turning point came from a process that physical stores and online had in common: product delivery. That was where we could act proactively and reach a large base of customers.
The question shifted from “how do we serve better” to “how do we anticipate problems”.
From insight to tangible
We prioritized paths by time, risk and effort, always deciding together with the team. For the MVP, we chose a customer pain point that was feasible to build and had low impact on operations. That let us test the idea with controlled risk before scaling.
The hypothesis was validated when the customer responded to the contact and the case was closed, a sign that predictive service worked. From there, a platform was built to manage the cases flagged by the system.
Result
I left the company before following the final numbers, but the platform is still in use today, with gains for operations and room to expand.
What I would do differently
Today I would use AI in parts of the process, such as analyzing call center calls. We could reach the dissatisfaction patterns much faster and put more energy into synthesis and decisions with the team.