Designing my own job

As we went back and forth over two lines on the checkout page, ship it to your home or ship it to a pickup location, I realized nobody in the room could say when one was better than the other. There was a rough idea. There was no framework, no evidence we could point to, nothing that laid out the moments when a locker made more sense than a doorstep. We were about to guide a customer’s choice on a hunch.

I had been hired as a senior UX designer because that was the only open requisition. My manager knew me from a previous team and wanted a service designer, so I came in through the door that was open and joined a triad with a product manager and engineers, shipping features into the pickup experience. My first was a guidance system for customers arriving at an unfamiliar store, trying to find where the lockers were.

The checkout page showed how the work was set up. Pickup had its own predefined space on the page, and home delivery belonged to another team. Whatever we wanted a customer to understand had to fit inside our space. And the choice itself was bare: shipped to your home, or shipped to an access point, with nothing to say why one might suit you better. Say you were buying a gift and didn’t want the person it was for to find it on the porch. That was a reason to choose a locker, and we weren’t telling anyone.

I couldn’t write that message without knowing who it was for. Who used access points, what were they trying to get done, and what did they expect when they got there? The question moved from how to word an option to who our customers really were and what outcomes they were counting on.

My manager had brought me in for exactly this, so the change came naturally once it was obvious that the whole ecosystem had to be looked at together, so the features we built would have cohesion instead of arriving one at a time. I handed the guidance concept to a more junior designer and moved across to service design for the entire access point ecosystem. That meant giving up quick feature wins to slow down and understand our customers more deeply, so the long-term strategy would hold up. The way I break a service down comes from SERVQUAL: find the gaps between what customers expect and what they actually get, then ask which gap, closed, would have outsized impact.

I designed and ran the research: driver ride-alongs and route observation, contextual inquiry, moderated remote sessions, and co-creation workshops in person and online. My own fieldwork took me to Austin and to Madrid.

What I built from it was a service blueprint large enough to cover a wall. It connected our last-mile team to the access point team, and for the first time you could see where the two met and how the whole system fit together. In workshops, people gathered around it on the wall and traced their own part with a finger, and between workshops they explored it on our digital whiteboard. They told me they were blown away, and more than one said they wished they’d had it when they were onboarding. Until then, how these organizations connected, and what a customer went through from end to end, had been invisible. Now people could see it, and act on it.

And the more we filled it in, the bigger the world on it became. Customers told us about lockers placed where they didn’t feel safe, like a convenience store in a neighborhood they would rather not walk through, and how uneasy they felt going to collect a package they had already paid for. Receiving what you bought is one of the most important moments in the whole journey. It should be a moment of delight, and for them it was a moment of fear. Others had no official address at all, and described where a package should go by landmarks: next to the old dead tree, down this dirt road. In Madrid, I saw that what people expected of the service was shaped by culture, and would differ from market to market.

An access point turned out to serve nine archetypes, in two groups: people who use it out of necessity, and people who use it out of convenience. Some of them the organization didn’t know existed. I defined the Durable Customer Outcomes: each customer type, what it actually needs, and how the service should accommodate it.

The work kept widening the circle. I built relationships with the leadership of product management, engineering and UX research, with business development and marketing, and with leaders across the wider Last Mile organization beyond our access point team. Marketing and sales were coordinating where the business would grow, and I wanted product development pointed at the same opportunities, so what we built stayed cohesive. I kept reaching out, gathering data and research, and explaining what I was trying to do and why, with the blueprint and other artifacts in hand. I was invited into marketing meetings, then, on that same trip to Madrid, into meetings with business development and marketing, then onto tiger teams of individual contributors a level above me. The more I showed, the more rooms I was asked into, and the stories I told about where the service could go kept landing.

Those rooms showed me where the business was already planning to expand, so I aimed my proposal there: specific cities and countries, with local agencies running deep ethnographic and contextual research in each, so the access points there would be built around the people who would use them. When my director approved it, he told me it was the most important piece of research our department would do the next year.

The Durable Customer Outcomes were adopted as the organization’s shared definition of its customers. With my manager, I built on them to co-write a three-year proposal for the future of the design team within the access point organization, organized around each customer archetype, its outcomes, and how we could serve it best.

Neither proposal ran, though both were approved and had buy-in. Company-wide layoffs followed soon after and took most of the team before either could start. The checkout options from that first meeting were updated by a more junior designer; my part was the research and the strategy.

When I arrived, we were about to guide a customer’s choice on a hunch. By the end, the organization had a shared answer to who its customers were, a picture of the whole system on the wall, and an approved plan to go and meet the customers it had never seen.

The leverage point is different in every organization. At Amazon, buy-in runs on data and on the customer, so a proposal moves when the evidence ties a customer outcome to investments the business is already making and expanding. Framed that way I was improving the return on something already committed to rather than asking for something new, and it met far less resistance than the same argument had in organizations with a looser account of why they were spending what they spent.