Experience Insiders

The Problem

Every internal tool we shipped needed real users to test it, not a hallway hunch. Search, learning, time off, all of it. And every study started the same way, from zero, on the slowest part of the job: finding the right people.

For each project we went to the HR analytics team with a spec. Which population, global or one geography or some mix. Corporate, retail, or distribution-center employees. Which roles, and what level of seniority or leadership. How large a sample the read actually needed to be representative. Every request carried a business justification, and some carried more: retail employees during the holiday sell-in, roughly US Thanksgiving through Christmas, sat inside a blackout window, because the business does not interrupt the people carrying its most important weeks. Only once that list came back could we start reaching out at all.

Then we had to convince those people to take part, one email at a time. I had built enough marketing email by then to know the odds. On the Maryland Skipjacks, Pinnacle, and Discovery work, my open and click-through rates ran only a hair above an average ecommerce blast, and recruiting for research was no kinder. Add it up and a single project burned two to four weeks just to find and enlist the right sample, before one test question got asked. We paid that setup cost on every study and kept none of it.

What I Built

The design question was blunt: what if we recruited the testers once, and never paid that setup cost again? So we stopped recruiting from scratch. I founded Experience Insiders with the HR Service Innovation team to answer it, a standing panel of about 600 Nike employees who raised their hands, once, to help shape the tools they used every day. Recruit them once, draw on them for every study after. The target was a number, not a mood: take that two-to-four-week recruiting drag down to a week at most, for any study and any demographic. Hitting that was the business case.

Efficiency was only the first of three goals set at the outset. The panel also had to make the tools genuinely fit what employees needed and, the one that mattered most, build a real relationship between HR and the people it serves. As best any of us knew, it was the first time Nike HR had asked employees to help shape their own work environment instead of rolling something out at them. That relationship was a design goal from day one, which is why the biggest return later was as much human as operational.

One pool every team could pull from, with what came back synthesized on affinity boards into patterns a roadmap could act on. The site turned that into an ongoing capability the whole org could reach.

The same panel carried two kinds of study. Some were self-led, done on a member's own time: a pulse survey between meetings, a search A/B test, a High-5 click test. Others were moderated, run live with a facilitator, from NikeU navigation sessions to the global HR website's future-vision work. One roster, every format.

What made it repeatable was that we designed the panel as a service, not a mailing list, and decomposed it into touchpoints anyone on the team could run. A segmented master list let us pull the right sub-group for a study, say the people managers or a single country, instead of paging all 600 members. Branded comms set expectations and gave every activity a single clear way in. And a library of reusable research toolkits let any HR team stand up a study in days rather than build one from nothing.

Who Was In It

No one ever advertised Experience Insiders. By early 2021 it had grown to 636 volunteers on word of mouth alone, a program no one was required to join that people kept joining anyway. And it drew the right people. Nearly a third were people managers, and membership skewed to the working middle of the organization, not the executives at the top or the entry level at the bottom, but the people who live inside these tools all day. Tenure ran from brand-new hires to 35-year veterans, averaging about seven and a half years, and members came from 24 of the 39 countries where Nike had employees. That spread was the point: one standing pool could reach a store employee in one country and a corporate planner in another, without recruiting either separately.

As one member put it, the appeal was simple: "I like the chance to give feedback because I feel like an outsider at Nike a lot of times. I like getting to represent myself."

That is the program in a single sentence, said back by the person it was built for. The people a company most needs to hear from are usually the ones who feel like outsiders to it. Experience Insiders had one job: to bring them inside.

The DEE Hub

Behind the panel was a product, and behind the product was a prototype. Everything the panel actually did ran live in production: members signed up and took their studies on a real Drupal site. The DEE Hub was the other half, the part I prototyped in InVision, a design vision for the Digital Employee Experience portal, built to settle how the whole program should feel before engineering committed to it. Every Hub screen below is that prototype, not shipped product: a view of what any Nike employee would open to see what the team was building and to shape it before it reached everyone.

The Hub pulled employees into the panel, put working prototypes in their hands, and closed the loop by showing each volunteer what their feedback actually shipped, and when. Closing that loop was the point, not a nicety: a volunteer who watches their own comment turn into a shipped change is a volunteer who comes back.

DEE Labs was the beta shelf. Every week it listed prototypes and early releases open for feedback, each one tagged by what it needed, a test, a build, or a design review, and fronted by a single button to jump in.

The roadmap, named DEElivered, was the promise the program made to its volunteers: new value every two weeks, on real dates, each item marked delivered or still in the labs. The real work on it included a search experience spanning HR, the intranet, and NikeU, a Google Search integration, and a rework of the header navigation and information architecture.

Recruitment lived in the same flow. Any employee could read how the panel worked and join from the same page, in a single step.

Each screen above does one job. The full page does what none of the crops can: it puts the entire operating model in a single view, the two-week delivery cadence, the beta modules under test, and the open door into the panel, all holding together at once.

That was the whole idea: a closed beta employees were invited into, not a survey they were subjected to.

A Self-Led Activity, Up Close

This one study is the whole capability in miniature: how the panel turned an open design question into a decision that shipped. The roadmap inside the Hub listed a rework of the site's header navigation and information architecture. That line did not come from a hunch. Every DEE Labs study started from a stated hypothesis or a specific feedback ask, a field built into the test template before a single screen went out. Here is one of the studies behind the nav rework: the HR Navigation Exploration, a navigation test the panel ran on their own time, one member at a time from a survey link.

The setup was deliberately narrow. A member opened the prototype from a survey tab, got a single instruction, and clicked to begin. Then they went looking for one thing.

The study lived in what came next. We built the same task four times, each as a different navigation pattern, and filed each under a Nike color codename so we could compare them head to head. Same destination, four routes to it, one question: which structure gets a person to Employee Discounts with the least friction? Four people can defend four favorite menus in a room forever; running the same task four ways turns that from a matter of taste into a matter of evidence.

Cyan kept a dark top bar and opened the chosen section as a tall list down the left edge.

Magnolia hid everything behind a hamburger, then opened a full dark drawer that expanded section by section, and added two role-based sections, Manager and HR, that the others left out.

Volt put a search field in the header and expanded the chosen section inline, pushing the page down instead of covering it.

RacerBlue used the same header but dropped a full-width mega-menu that put the entire section on screen at once.

The evidence came back fast. Nineteen members ran the first pass, a deliberately small, directional read rather than a statistical verdict, and it was enough to point the direction. Two patterns pulled ahead: the left-edge list and the clean search-header that dropped the section down from the top, each getting about nine in ten of them, roughly seventeen of the nineteen, to Employee Discounts without trouble. Those two sat within noise of each other, so the read was a direction, not a winner crowned by a single click. The full-page hamburger overlay landed last, the only route a quarter of members called outright difficult, dragged down by an extra click and a hamburger not everyone expected. Asked to rank the set, they put the search-header drop-down first and the left-edge list a close second.

But the sharpest finding had nothing to do with which of the four menus won. It cut across all of them. Members called the second-level menu "too much information," "very list based and wordy," overwhelming no matter how it opened. And they kept reaching for Employee Discounts under Benefits, not wherever the taxonomy had actually filed it.

That was the moment the study stopped being a menu test. The question stopped being which of the four patterns won and became whether the information architecture itself was filed wrong. So the readout did more than pick a menu: it said to thin what each menu held and move things to where people expected to find them. That is what pointed to the header-navigation and information-architecture rework on the roadmap, not a preference in a room. The navigation it pointed to shipped on the global HR website, where the winning top-bar structure and its full readout live.

Impact

The operational win came first, and it answered the problem we opened on: the recruiting drag that used to front every study, two to four weeks of speccing a population, waiting on a list, and emailing strangers, dropped to a week or less. One shared panel meant no team had to start a new tester list for every launch. In a single month that one pool could carry half a dozen studies at once, a pulse survey and a search A/B test running the same week as moderated NikeU and HR-website sessions, each stood up in days off the shared roster instead of built from scratch. The same pool also ran enterprise search, and what it showed went onto dated roadmaps: a launch got pressure-tested before it shipped, on the testers' own time, and the drag that used to front every study was gone.

The bigger return was human, which is the day-one goal finally paying out, not a lucky by-product. The panel's members got real say over the tools they were handed every day, and the people who tested a release became the ones who vouched for it. This is where closing the feedback loop compounded: members who watched their own comments ship came back, and the program kept growing on that goodwill. The response ran deep enough to carry a full second version, the Experience Insiders 2.0 site, into final QA by early 2021. As another member put it, "At Nike we see change all the time, and rarely do we know why. Having that insight makes us all more supportive of the change."

I ran Experience Insiders under HR Service Innovation for four years, from its earliest days. In September 2022 I handed it to Nike's HR Sr. Director of Employee Experience as a standing capability, not a project that ended with me, and moved to Amsterdam for the EMEA role Nike relocated me for.

What Transfers

Strip out the Nike specifics and the operating move is what is left, the part that carries to any company. Recruit the people a change will land on, once, and give them a real stake in it, and adoption stops being something you push and becomes something they carry. Decide from evidence a nineteen-person test can produce, not from the loudest voice in the room. And build the capability with an owner from the start, so it outlives you, the way this one outlived my handoff.

That is still how I work. The bet underneath Experience Insiders, that the people closest to a problem will beat any amount of top-down research once you treat them as partners instead of subjects, is the same one I am making now.

Experience Insiders was built and run with Sergio Slansky and Sarah Sparks.