Rebranding without the brand.

In 2023, Aramark spun off its uniforms and workplace supplies business as a separate public company called Vestis. It got a new name, a new logo and a new domain. The old uniforms site had years of search history behind it. A big share of its organic traffic (visits from unpaid search results) came from people searching for the Aramark name.

That name was staying with Aramark. The site had to move to Vestis and leave the name behind.

A 301 redirect tells Google a page has moved for good, and it passes most of the old page’s ranking signals to the new one. That takes care of the pages. It does nothing for everyone who keeps typing “Aramark uniforms” into a search bar.

Some loss was certain. The work came down to three things:

Before any work on the new site started, I crawled the old one with Screaming Frog to get a full inventory. Every URL got its own row in a spreadsheet, scored on:

The matrix did two jobs. It showed which pages mattered most, so they were mapped and protected first. It also showed how much of the site’s traffic depended on the Aramark name, and that’s where the forecast came from.

I used the same approach, with leads added as a metric, on the Wasabi replatform.

Using the matrix, I estimated how much branded search traffic the new domain would lose. Then I walked stakeholders through that forecast before launch. Nobody likes hearing traffic will go down. They like it a lot less when it’s a surprise.

The migration was a chance to fix the structure, not just copy it to a new domain. Before mapping a single redirect, I designed a new information architecture for Vestis: how pages are grouped, how they link to each other, and how someone gets from a broad topic to the exact service they need.

The site was organized two ways: by product delivery service, and by industry. A restaurant manager and a facilities buyer can each start from the path that fits how they think and still reach the same pages.

Settling the structure first meant every later decision had a place to land. New content had a home, and every redirect had a clear destination.

Branded search was going to shrink no matter what. So the long-term fix was to earn traffic from people who had never heard of Aramark or Vestis, but were searching for what the company actually does.

I ran non-branded keyword research to find those searches. That turned into about 40 new pieces of content, grouped into topic clusters. A cluster is a main page on a subject, with supporting pages around it that all link together, which helps search engines see the site as an authority on that topic.

Every cluster was built into the new IA, not bolted on afterward. The clusters went live with the new site in February 2024.

With the structure and content in place, I mapped the old site onto the new one. Two kinds of redirects did the work:

A redirect only works if the destination deserves it. A page that ranked for a specific search had to land on a page that answered the same search. So on-page alignment ran alongside the mapping: titles, headings and content on each priority destination were adjusted to match what the old page had earned rankings for.

Vestis serves customers from physical locations, and a lot of local searches never reach a website. People see the listing in Google or Bing and call or get directions from there. Those listings still carried the old brand.

We built nearly 350 unique landing pages, each optimized for its local market, and paired every one with an updated Google Business Profile and Bing Places listing. The content was generated programmatically, which is the only realistic way to do 350 pages well without spending a year on it. Local search showed Vestis, with accurate details, where people were already looking.

Every redirect was tested in a staging environment before launch. Then we tested them all again in production. Staging is a rehearsal, and plenty of things work in rehearsal and break on opening night. We confirmed every 301 was in place on the live site.

After launch, I monitored traffic against the forecast and coordinated the old and new Google Search Console properties, so Google had a clear picture of where everything had moved.

Core Web Vitals are Google’s measures of how fast a page loads, how quickly it responds, and how much it jumps around while loading. After launch, the new site’s scores sat in the “needs improvement” range. With branded traffic down as forecast, the client was anxious and wanted everything green.

There were real fixes to make. I found oversized images and images that weren’t lazy-loading, meaning they loaded before anyone scrolled to them. The team found that the chat widget and heatmap tool were slowing pages down, and we optimized how they loaded.

The bigger job was keeping the focus in the right place. Core Web Vitals mostly act as a tiebreaker in rankings. The traffic dip came from retiring the Aramark domain, which the forecast had predicted. So the fixes went in, and the main effort stayed on recovery: the content, the redirects and the local pages.