Most redesigns go wrong in the same quiet way. The new site looks better, ships on time, and then traffic falls off a cliff in week two because nobody mapped the old URLs to the new ones. A disciplined website redesign process treats the visible work (the design, the build) as the middle of the job, not the whole of it. The risk lives at the edges: the audit you skip and the redirects you rush.
What follows is a step-by-step guide to running a redesign that improves the site without sacrificing the rankings and traffic you already earned. It is opinionated, because every step exists to prevent a specific failure we have watched teams walk into.
Start with an audit, not a moodboard
The temptation is to open a design tool and start exploring. Resist it. The first artifact of a redesign is a clear-eyed inventory of what you have, what works, and what is quietly load-bearing.
A useful audit covers three layers at once:
- Analytics. Which pages drive traffic, conversions, and revenue? Pull at least twelve months so seasonality does not fool you. The pages that matter are rarely the ones the team is proud of.
- Search. Which URLs rank, for which queries, and with how much authority? These are the pages you protect at all costs.
- Content and tech. A full crawl gives you every URL, redirect, orphan page, and broken link you are about to inherit.
A redesign without an audit is a renovation where nobody checked which walls hold up the roof.
The audit produces the most important deliverable of the project: a list of your highest-value URLs. Everything downstream, including the architecture, the content plan, and the redirect map, references this list.
Define goals you can actually measure
“Make it modern” is not a goal. It is a feeling, and feelings do not survive contact with a deadline. Before any structure or design work, write down what the redesign should change and how you will know it worked.
Good redesign goals are specific and few: increase demo requests from organic traffic, hold rankings on your top twenty commercial queries through the migration, reduce the page weight costing you on mobile. Name the metric for each goal now, while it is cheap to argue about. A goal without a metric is a wish, and a redesign full of wishes pleases everyone at kickoff and no one at launch.
Information architecture comes before any pixels
With the audit done and goals set, the next move is structure, not layout. Information architecture decides what pages exist, what they are called, and how visitors move between them. It is the cheapest thing to change and the most expensive to get wrong.
This is also where most redesign traffic loss is either prevented or guaranteed. When you reorganize a site, URLs shift, pages merge, and sections disappear. Every one of those moves is a redirect you will owe later. We go deep on this in our guide to website information architecture, but the redesign-specific rule is simple: design the new structure with the old URLs in hand, so you always know what maps where.
Keep the top-level navigation tight, group by what visitors are trying to do rather than by your org chart, and resist clever section names that crawlers and first-time visitors will both fail to parse.
Sort the content before you design around it
Content is the part everyone underestimates and the part that delays the most launches. You cannot design real pages around placeholder text, and you cannot migrate cleanly if half of it is unwritten when the build starts. Work through your inventory and tag every page with a decision:
- Keep. Migrate as-is, or with light edits.
- Improve. Rewrite or restructure, usually your high-value pages that have grown stale.
- Merge. Combine thin or overlapping pages into one stronger page, and remember that every merge is a redirect.
- Cut. Remove pages with no traffic, no links, and no purpose. Cutting is healthy; just track what you cut so you can redirect it.
- Create. New pages the strategy demands that do not exist yet.
Doing this early turns content from the thing that blocks launch into a parallel workstream. Editorial quality is not a finishing touch; it carries the brand, which is why we treat content as a first-class part of the build rather than a step bolted on at the end.
Design against the structure, not the other way around
Now design. With architecture settled and content sorted, design becomes a problem of expressing decisions you have already made rather than improvising meaning from aesthetics. Start with the page types that do the heaviest lifting, usually the homepage, the primary service or product pages, and the conversion paths. Design the templates, not every individual page, so the system stays coherent and the build stays sane. Accessibility is a decision you make in these layouts, not a checklist bolted on before launch; a redesign is the rare moment to fix it at the source instead of patching it later.
Build for speed and for the people who maintain it
The build is where a good redesign either compounds its advantages or quietly throws them away. Two things matter most: how fast the site is, and how maintainable it stays.
Speed is not a nice-to-have. It is a ranking factor, a conversion factor, and the first thing a visitor judges you on before they read a word. The real cost of a slow website is paid in bounced visitors and lost rankings, so performance belongs in the build from the first commit, not a frantic pass the week before launch.
Maintainability is the other half. The redesign should leave your team with a site they can run: content in a system editors can use, a codebase a new developer could pick up, no proprietary lock-in. This is the standard we hold ourselves to when we build websites and apps: the team that scopes the work builds it, so what gets handed over is something the client owns outright, not a black box.
The redirect map: where redesigns live or die
Here is the step that quietly decides whether your traffic survives. When URLs change, search engines and the links pointing at you need to follow the move. A redirect map is the spreadsheet pairing every old URL with its new home.
Build it methodically:
- One row per old URL. Pull the full list from your crawl and analytics, not just the pages you remember.
- Map every high-value URL to its closest equivalent. A merged page redirects to the page it merged into; a renamed page to its new path.
- Use permanent (301) redirects so authority transfers and the change is treated as final.
- Never bulk-redirect everything to the homepage. Search engines read that as a soft removal and you lose the ranking signal. Map to the most relevant page, every time.
- Avoid redirect chains. Old URL to new URL in one hop, not three.
The redirect map is the single most underfunded artifact in a redesign and the single most common reason traffic drops after one.
Test the map in staging before launch. A few minutes verifying that your top fifty URLs resolve correctly will save a frantic week of recovery later.
Launch deliberately, not on a Friday
Launch is a procedure, not an event. Treat it like one and most of the drama disappears.
Before you flip the switch, run the pre-launch pass: confirm the redirect map is live, check that analytics and conversion tracking fire on the new pages, validate your sitemap and robots.txt, and make sure nothing is still set to noindex from staging. That last one has sunk more launches than any design flaw.
The moment you go live, submit the new sitemap in Search Console, spot-check your highest-value redirects in production, and watch your error logs for 404s. Launch when you have time to monitor, early in the week and early in the day, never at 5pm Friday.
Measure against the goals you set
A redesign is not finished at launch. It is finished when you have confirmed it did what you set out to do, which is why those measurable goals from earlier matter now.
Expect a brief dip as search engines recrawl the site; a short wobble is normal, a sustained decline is a signal. Watch rankings on your priority queries, crawl errors in Search Console, Core Web Vitals in the field, and the conversion metrics you named at the start. Watch AI assistants too. Increasingly, getting cited by tools like ChatGPT and Perplexity depends on the same clean structure and legible content that protects your search rankings.
If something slipped, the audit and the redirect map are your way back. Nine times out of ten, a post-launch drop traces to a missed redirect or an accidental noindex, both fast to fix once you know to look.
How Strynal approaches a redesign
We run the website redesign process as one continuous arc, not a relay of handoffs. Every engagement starts on a blank page (no templates, no recycled layout from the last client) because we work with people solving uncommon problems, and the audit almost always reveals the real opportunity is somewhere no one expected.
Strategy, brand, and build sit under one roof, so the architecture, content, design, and redirect map stay in agreement instead of drifting apart between vendors. The senior people who scope the work are the ones who ship it, which is why the unglamorous steps (the audit, the redirect map, the launch checklist) actually get done. As the in-house studio for Global Digital Platforms, we have learned those quiet steps separate a redesign that grows traffic from one that loses it.
If you are planning a redesign and what keeps you up at night is losing the rankings you already have, tell us what you’re working with. We will start where every good redesign starts: with a map of what you cannot afford to break.