Most sitemaps grow by accident. A page gets added for a campaign, another for a new product line, a third because someone senior asked for it. Five years later the navigation has eleven items and nobody’s sure which pages still matter. A redesign is the moment to fix that, and we try to do it in a single two-hour workshop.
Before the room
The workshop only works if the homework is done first. In the week before, we gather three things:
A full crawl of the current site, with every URL, its title and the template it uses.
Twelve months of analytics: entrances, conversions and search traffic for every page.
The client’s written brief, especially who the site is for and what those visitors should do.
From the crawl, we print a card for every page that’s had meaningful traffic or links in the past year. For a typical business site, that’s somewhere between 40 and 120 cards. Pages with no visits and no links go straight onto a separate list for redirects. We rarely need to talk about them.
We also ask the sales team which five pages they send to prospects most often. Their answer rarely matches the analytics, and those pages almost always matter.
While we’re at it, we note every page with a form, booking widget or other connection to the rest of the business. A contact form that feeds a CRM is easy to lose in a rebuild, and nobody notices until the leads stop arriving.
The four rules
We open the workshop by agreeing on four rules. They sound strict. They save hours.
Every page needs a visitor. If nobody can say who it’s for, it doesn’t make the cut.
Every page needs an owner. One person in the business who’ll keep it accurate.
No more than six items in the main navigation. Everything else lives one level down, in the footer or inside the content.
Names come from the customer’s vocabulary, not the org chart.
Rule four moves more cards than the other three combined. Customers don’t search for your department names.
Sorting the wall
We stick the cards along one long wall in three columns: keep, merge and remove. The team sorts them quickly and out loud, with the analytics printed on each card, so opinions have to meet numbers. A page someone loves that had forty visits last year can still stay, but it becomes a different conversation.
Merging is where most of the value is. Five thin service pages often become one strong page with sections. Three overlapping About pages become one. Every merge gets an arrow on the wall showing which old URLs will point where, and those arrows become the redirect map later.
Grouping by intent
Once the keep pile is settled, we group the remaining cards by what the visitor is trying to do, not by topic. Typical groups are “understand what you do,” “see proof,” “check the details” and “get in touch.” Those groups become candidates for the main navigation, and the labels get rewritten in the words customers actually use. Search terms from Google Search Console settle most naming debates.
Then we walk the two or three key visitors from the brief through the new structure, one at a time: from the page they’d most likely land on to the action we want them to take. If a journey takes more than three clicks, or passes through a page with no clear next step, we keep moving cards until it doesn’t.
What leaves the room
At the end of the two hours, we photograph the wall. Within a day, we send back three documents:
The new sitemap, with each page’s visitor, owner and template type.
A redirect map from every old URL to its new home.
A content list of everything to write or rewrite, with an owner and a date for each.
The sitemap becomes the backbone of the design phase. Each template type is designed once and reused, which is why a well-planned sitemap usually makes the build cheaper as well as clearer. And because the redirect map exists from week one, protecting your search rankings at launch is a checklist, not a scramble.
Try it on your own site
You can run a smaller version without us. Export your pages from analytics, print the top fifty, and sort them with two colleagues against the four rules. Even if you’re not planning a redesign, you’ll find pages to merge, labels to change and at least one navigation item nobody can explain.
