Website Redesign vs New Website: How to Make the Right Decision in 2026

Compare the bottleneck, not the prices. And there are usually four options, not two.

SHORT ANSWER

Redesign or New Website: How to Decide

Redesign when the foundation of the site is healthy and the problem is its look, its message or the structure of its pages. Build a new website when the technology itself has become the limit: every change needs a workaround, the system is no longer supported, or it cannot do what the business needs. Two more options sit between these and come up less often: fixing specific places on the live site, and replacing the code while the look stays.

Most businesses compare only the last two, and only by price. That comparison works badly, because a project's price says little about the state of the site three years later. This article is about finding what has actually become the bottleneck, and about what we chose on our own projects.

THE QUESTION

What Problem Are You Actually Solving?

The conversation usually opens with "should we redesign our website?" A more useful question is what business problem we are trying to solve. More qualified leads, more trust, support for a new line of business, less manual work inside the company. A redesign may be the answer to any of these, and it may not. A website is often the solution, and it is surprisingly rarely the problem itself.

If visitors do not arrive, a new design will not change that: search, targeting and content do. If they arrive and do not understand what you sell, structure and copy help. If they understand and still do not reach out, the cause is often a form, a button or missing proof at the moment of decision, and that can be fixed without a big project. These three layers are covered in Why Website Redesigns Don't Bring More Leads, and checking them against real behaviour, with session recordings and heatmaps, in Do You Need a Website Redesign? Start With a Usability Audit.

It helps to think of a site as a commercial building. If the foundation is solid, the structure is reliable and the layout still works, renovation usually makes financial sense. If every improvement uncovers another structural issue, rebuilding eventually becomes the less expensive option. The point is not what looks old, but what already gets in the way of work. Often the real need turns out to be a new structure and not a new look.

Looking old and getting in the way are two different diagnoses.

REDESIGN

When a Redesign Is Enough

A redesign usually pays off when the site has a healthy foundation. Typical signs:

• the site is visible in search, and those pages bring people in;
• the CMS and admin panel are reliable, and editors manage on their own;
• the page structure is logical and the content is current;
• speed is acceptable and technical failures are rare;
• the main issue is the brand, the visual language or ease of use.

In that case a redesign renews what the customer sees and keeps what took years to build: search positions, content, integrations. That does not mean inquiries will rise on their own. A redesign whose brief describes only the look tends to deliver only the look. If the goal is measured in inquiries, it belongs in the brief, and the baseline numbers need recording before work starts.

NEW WEBSITE

When You Need a New Website

Sometimes the website itself becomes the limit. Typical warning signs:

• the platform is no longer supported, and updating it is difficult or impossible;
• the site is slow, Core Web Vitals are poor, and fixes do not help;
• the mobile experience is weak because the layout was never written for phones;
• technical failures repeat, and fixing one breaks another;
• editors cannot manage content without a developer;
• there are security concerns;
• the business has grown, and new features run into the old system.

If every improvement needs another workaround, plugin or compromise, you are no longer investing in growth but in keeping an old system alive. One of our clients runs a large custom system on PHP 5.3, whose security support ended in 2014. It works, but its connections rest on fixes added one at a time over many years, and an upgrade would carry all of them into the new version. So we signed a contract for a full rebuild.

It also happens that the request is to improve the live site and a new one turns out to be simpler. That was the case with SHTAYER. The store had been running since 2020, and the company first came back to us with refinements to the site and its functions. Then the question of the identity came up, and B2B and B2C also had to be joined: almost every product carries its own logic of sizes and materials, and private customers, interior designers and hotels buy it at different prices and on different terms. The platform the store launched on ran out of room for that logic, so the store is now moving to Laravel, and the refinements to the live site are enough to keep working until the new one is ready.

It also works the other way. The Institute of Trichology was ready to pay for a new website, and we started with the technical debt instead: we updated PHP and OpenCart and kept the existing design. What came of it is in our piece on websites for specialty clinics.

That decision has a price. In 2000 Joel Spolsky called rewriting code from scratch the single worst strategic mistake a software company can make, and his reason still holds: old code contains hundreds of fixes nobody remembers. A new website starts clean only in its code. What the old site knows (which pages matter, which paths bring inquiries, what people search for) has to be extracted first, or the old mistakes return in a new form. A large system can also be replaced in parts, next to the working one: Martin Fowler calls this Strangler Fig, and in it both investment and returns come gradually and visibly.

If a new site really is the answer, the full cycle from strategy to launch is on our website development page, and the design work on its own on website design.

NEW CODE

The Third Option: New Code, Same Look

This option is offered least, though on our projects it comes up very often. The interface suits people, and underneath it sits an old framework version, unmaintainable CSS or an admin panel nobody enjoys. Visitors do not notice the replacement: the site simply becomes faster and more stable.

The best example is close at hand: the site you are reading this on. In early October 2026 we moved it from Laravel 7 to Laravel 13 and from PHP 7.4 to PHP 8.4, with about three days of development, one day of acceptance and five minutes of downtime, described in detail here. Nothing changed on the surface. We began the redesign afterwards, on the upgraded system, and released it one page type at a time: contacts, about, blog, articles, portfolio, services. Addresses and copy stayed as they were, and across all 513 article URLs we compared the old and new versions: titles, meta tags, canonical, hreflang, H1, structured data, content, links and image alts.

The order matters more than the caution. Google's guide to site moves says plainly to change one thing at a time. If the engine, design, structure and addresses all change on one day and rankings fall, nobody can tell what caused it.

We make these calls on more than our own site. When we took over projects from other contractors, an upgrade proved the right answer in two cases out of three, and a full rebuild in one. And we decided against redesigning our own partner studio, DeZZign, altogether: the site works, and the data showed no problem a redesign would solve.

The most useful part of a project can stay invisible to visitors.

MONEY

What It Costs and What to Count

Here are our own prices as of November 2026, as a reference point. Other markets and contractors will differ, and the full list is on our pricing page.

The first figure in an estimate is not the most important one, because the real cost of a website builds up over three to five years, and that is what to count:

A redesign may cost less today and a new website less over five years. It also happens the other way round: an expensive new site repeats the old mistakes and asks for money again three years later. What surfaces inside such projects is in our article on hidden technical debt.

SEARCH

What Happens to Your Rankings

A redesign does not have to hurt SEO, and a new website does not have to lose it. The damage comes from specific things: changed addresses without 301 redirects, deleted copy that "did not fit the new design", a slower mobile version. Google warns that any significant change can bring temporary ranking fluctuations, and that a medium-sized site usually needs a few weeks or more before new addresses replace the old ones in results, and a large one longer.

So before work begins, list the pages that bring people from search today and decide the fate of each. Keep the address wherever possible. Before launch, check that forms send email, analytics counts inquiries and redirects lead to the right pages. The full list is in our website migration checklist.

QUESTIONS

Seven Questions Before You Decide

Before making a significant investment, answer these with your partner or leadership team. Go by today's reality and not by assumptions.

1. Does your website still represent the company you've become? Businesses often change faster than their sites: new services, better positioning, higher-value clients, while the screen still shows the company from five years ago.

2. Is the site supporting your sales process or slowing it down? If your team keeps answering the same questions, prospects hesitate before contacting you and visitors leave without acting, the site no longer supports the business.

3. Are you investing in improvements or constantly fixing problems? Improving a healthy platform creates value. Continuously repairing an unhealthy one creates expenses.

4. If you launched your company today, would you build this same website again? If the honest answer is no, the decision is probably no longer about design, but about whether this site still deserves to remain the foundation.

5. Where do you expect the business to be in three years? The site has to carry the company you are becoming. Additional services, automation, CRM and new markets are worth naming in advance.

6. How will you measure success twelve months after launch? Without an answer, nobody will be able to say a year later whether the project worked.

7. What happens if you change nothing? This question can bring more clarity than a dozen agency proposals.

MISTAKES

Where Businesses Most Often Go Wrong

Choosing on price alone. The cheapest option often becomes the most expensive after several years of fixes and compromises.

Judging the look and not the result. A beautiful site that brings no inquiries is an expensive portfolio piece. Design matters, and the test of it is what people do.

Ignoring growth. A site built only for today's needs soon has to be built again if it cannot take new services, landing pages, CRM integration and campaigns.

Leaving SEO for later. It shapes architecture, content structure, page hierarchy, links and speed. It starts before the first mockup.

Changing everything at once. If something goes wrong, nobody will know what did it. New code, then design, then structure is usually safer than all of it in one day.

Solving the wrong problem. Weak positioning, a lack of trust, poor leads or too little traffic are not solved by a new website automatically. This is probably the most expensive mistake of all.

FAQ

Sources

• Google Search Central: site moves with URL changes
• Joel Spolsky: Things You Should Never Do, Part I
• Martin Fowler: Strangler Fig