Before You Redesign Your Website, Figure Out What’s Actually Broken

“We just need the website to look more modern.”

It sounds reasonable. It is also one of the most expensive sentences in a redesign project.

A website can look dated and still have a deeper problem. It can also look beautiful and fail commercially. The visual layer is the easiest part to notice because everyone can see it. Strategy, information architecture, search equity, content gaps, performance, analytics, and internal workflow are less visible, but they are often where the real cost sits.

A redesign should not begin with a moodboard. It should begin with a diagnosis.

The redesign brief is usually written too early

Many businesses decide on the solution before they understand the problem. The brief says “new website,” then the project immediately moves into references, page counts, features, and design direction.

That sequence creates a predictable risk: the team rebuilds the current site's assumptions in a cleaner visual system.

If the existing navigation reflects an outdated service model, the new navigation will still be wrong. If the homepage speaks to three different audiences at once, a more polished homepage will still be vague. If high-performing search pages are deleted because nobody knew they mattered, the new site can launch looking better while performing worse.

Before design, we use six lenses: strategy, message, journey, search, performance, and operations.

1. Strategy: what job is the website actually being hired to do?

“Generate leads” is not enough. A serious website needs a more precise commercial job.

Is it meant to qualify larger clients? Reduce dependence on founder-led sales? Make a complex offer easier to understand? Support geographic expansion? Improve recruitment? Give sales teams better proof? Shift the perception of the company from vendor to strategic partner?

Those goals produce different websites.

A site designed to increase qualified enquiries may intentionally discourage low-fit leads. A site supporting a long B2B sales cycle needs depth, evidence, and internal-shareability. A site for a premium service business may need fewer conversion points, but stronger ones.

If the redesign cannot name the business behaviour it is meant to change, the project is still too vague.

2. Message: can people understand the business without decoding it?

Website projects often expose brand problems. Teams discover that the reason the homepage is difficult to write is not a copywriting failure; the company has never made a clear decision about positioning.

Audit the current site page by page. Where does it use category language buyers actually understand? Where does it rely on internal terminology? Which claims could belong to any competitor? Which proof points are buried? Where are important objections unanswered?

A good redesign creates a content hierarchy before a visual hierarchy. It knows what the reader must understand first, what they need to believe next, and what evidence earns the right to ask for action.

3. Journey: where does the site make the visitor work too hard?

Information architecture is not a sitemap exercise. It is a decision-path exercise.

Map the main visitor types and the questions each one brings. A new prospect may need to understand fit. A returning prospect may want proof. A referral may already trust the company and simply need a clear path to contact. A potential hire is judging culture. A partner is looking for credibility.

Then audit the current journeys. Look for dead ends, repeated information, pages that exist only because they always have, and calls to action that ask for commitment before enough confidence has been built.

The goal is not to minimize clicks at all costs. The goal is to remove unnecessary uncertainty.

4. Search: protect what already works before you improve what does not

Redesigns can accidentally damage years of search equity. URLs change, pages disappear, internal links break, titles are rewritten, content is consolidated, and important landing pages are removed because they were not visually impressive.

Before touching structure, export the current URL inventory and performance data. Identify pages receiving organic traffic, backlinks, impressions, conversions, and branded demand. Decide which pages stay, which merge, which improve, and which genuinely deserve to disappear.

If URLs change, create a redirect map before launch. Google's own site-migration guidance emphasizes mapping old and new URLs and using appropriate redirects when content moves. This is not glamorous work, but neither is explaining why qualified organic traffic vanished two weeks after launch.

5. Performance: speed is part of the experience, not a technical appendix

A visual redesign can easily become heavier than the site it replaces. Large video, oversized images, multiple font files, animation libraries, third-party scripts, and unnecessary interactions can turn “premium” into sluggish.

Performance should be designed in, not optimized as a final sprint.

Google's current Core Web Vitals framework focuses on loading performance, responsiveness, and visual stability through LCP, INP, and CLS. The useful part is not chasing a score for its own sake. It is treating speed and stability as real interface qualities. A page that jumps while loading, responds slowly to input, or keeps the main content waiting is asking the visitor to pay a usability tax.

Set performance budgets early. Decide which motion is important enough to keep. Compress and size media intentionally. Avoid shipping functionality nobody uses.

6. Operations: can the team actually run the site after launch?

Some websites fail slowly, after the agency leaves.

The CMS is awkward, so nobody publishes. Every minor update needs a developer. Marketing cannot create landing pages without breaking layouts. Reusable components were never planned. Analytics are inconsistent. Forms are not routed correctly. The site becomes a polished object rather than a working business system.

Before redesigning, ask who will own the website, what they need to update, how often content changes, which pages require reusable structures, and what integrations are genuinely necessary. Platform decisions should follow operational reality, not trend loyalty.

A composite case: when the “design problem” was really three problems

Imagine a specialist service firm preparing for a major redesign. The current site is five years old. Leadership describes it as “dated” and asks for a premium visual upgrade.

The audit finds three more important issues.

First, the company's revenue has shifted toward a higher-value service, but the website still leads with the legacy offer. Second, organic search traffic is concentrated on a handful of educational pages that were not included in the proposed sitemap. Third, the marketing team cannot create new campaign pages without duplicating old sections manually.

If the project had moved directly into design, all three issues would have survived, and the strongest search assets might have been deleted in the process.

The better redesign brief becomes clear: reposition the offer hierarchy, preserve and improve high-value search content, build a reusable CMS and component system, then create a visual language that supports the new market position.

That is a redesign worth paying for because the scope is connected to business behaviour.

Accessibility belongs in the redesign brief

Accessibility should not arrive as a checklist after design approval. It affects contrast, type size, focus states, keyboard interaction, target sizes, motion, forms, labels, content structure, and media.

WCAG 2.2 provides the current international reference point. You do not need to make the site visually conservative to make it accessible. You do need to stop treating accessibility and high-end design as opposing ideas.

Good interaction design makes a site easier to use for more people. That is not compliance theatre. It is product quality.

Build the redesign brief from evidence

A useful pre-redesign brief should answer:

• What commercial outcome must the new site improve?
• Which audiences matter most?
• What positioning or messaging must change?
• Which pages already perform and must be protected?
• Which user journeys currently create friction?
• What performance standards are non-negotiable?
• What accessibility requirements guide design and development?
• What does the internal team need to manage without outside help?
• How will success be measured after launch?

Only then should the conversation turn to visual direction, motion, component systems, and creative references.

A redesign is a business intervention with a visual output

The strongest websites are not the ones with the most fashionable interface. They are the ones where strategy, message, content, interaction, technology, and operations reinforce each other.

That is why the right question before a redesign is not “What should the new site look like?”

It is “What is the current site preventing the business from doing well?”

Answer that honestly, and the design work becomes much more interesting. More importantly, it becomes much harder to waste.