Change management case study · Walmart eCommerce · 2019
Updating Walmart's Search Algorithm
A three-month, multi-channel internal communications campaign that drove adoption of updated customer search algorithms across a 2,400-person merchandising organization, producing $8M in revenue lift within six weeks of launch.
- Company
- Walmart eCommerce
- Audience
- 2,400 merchandisers and managers
- Duration
- 3 months
- My role
- Strategy and full execution
The problem
If you searched 'frozen dress' on Walmart.com, you might get a shoe, a t-shirt, or nothing at all. One in four searches on the site was returning unrelated products or no results, eroding trust with both customers and suppliers. The search team had diagnosed the issue and built a fix that required thousands of merchandisers to manually review and rework their product queries alongside their daily workload. My communications needed to explain why the work mattered and help teams make time for it.
The strategic decision
Once merchandisers understood how the search issues affected customers, they understood why the work mattered. The challenge was getting them to prioritize it alongside everything else they were responsible for.
I worked with managers to help make time for the updates and introduced a team competition to encourage participation. I also brought in someone from the search team who had previously been a merchandiser. He knew the challenges they faced and could explain the changes and answer questions from their perspective.
What I built
The campaign ran for three months and included targeted emails, weekly newsletters, in-person Q&A sessions, manager briefings, and a competition between departments. We consistently explained the search problem, what merchandisers needed to do, and how their work would help improve the customer experience.
We shared progress and results throughout the campaign so teams could see the impact of their work. A leaderboard tracked each department's progress, and we publicly recognized the winning teams.
What I'd do differently
I'd build in a measurement layer to track behavior change following the campaign. Having proof points that the new querying behavior stuck would have made the case for future campaigns much stronger.
The broader lesson
Showing merchandisers the customer problem helped explain why the query reviews mattered. Manager briefings and a visible team leaderboard gave them support and a way to follow progress.