The traffic dip has become almost a standard clause in how agencies talk about website redesigns. There will be a period of adjustment. Rankings may shift temporarily before they recover. Google needs time to reindex the new site. All of this is delivered as unavoidable fact, and most businesses accept it. Some of it is true. Most of what causes the traffic dip is not Google needing time. It is the result of specific, preventable decisions made during the redesign process. Agencies that understand this deliver redesigns where rankings hold. Agencies that do not understand it, or do not prioritise it, deliver redesigns followed by months of recovery.
Understanding what actually causes post-redesign traffic loss is the starting point for avoiding it.
The ranking audit that almost never happens before design begins
Most website redesign briefs include some mention of SEO. The mention is almost always about optimising the new site: better metadata, improved page speed, content improvements. What the brief rarely includes is a systematic audit of what is already ranking on the existing site and why.
This matters because a redesign that does not account for existing ranking assets will routinely destroy them in the process of building something better. A page that ranks on the first page for a valuable keyword represents accumulated effort: backlinks, content depth, dwell time, internal linking. If that page disappears in the redesign, or its URL changes without a proper redirect, or its content gets thinned out in favour of cleaner presentation, the ranking disappears with it.
The correct starting point for any redesign with SEO implications is a ranking audit before design begins. Which pages currently hold ranking positions? What keywords are they ranking for? What is the monthly traffic and commercial value of each ranking page? What is the content and on-page structure that appears to be supporting those rankings? None of this is complex analysis. It requires Search Console access and a few hours of focused work. It also directly determines which pages are too important to redesign without careful attention and which can be changed freely.
Agencies that skip this audit make design decisions that feel right visually but cut across SEO considerations that have taken years to build. The most common version is consolidating multiple service pages into one overview page because it produces a cleaner navigation structure. From a design standpoint this is reasonable. From an SEO standpoint it can eliminate three or four individually ranking pages in exchange for one page that ranks for nothing initially.
URL changes and the redirect tax
Every URL change in a redesign is an SEO risk. This is not because 301 redirects do not work. They do. It is because redirect chains accumulate, redirects get missed, and even a correctly implemented redirect transfers somewhere between 90 and 99 percent of the original page's ranking value rather than 100 percent.
For a business with a handful of high-value pages, the impact of URL changes can be contained. For a business with a large content collection, a CMS migration or a restructured information architecture can involve hundreds of URL changes. Each one is a small leak in ranking equity. At sufficient scale, the cumulative effect is material.
The practical approach is to preserve existing URL structures wherever possible. If a URL has been live for more than two years, has accumulated inbound links and holds a ranking position, the default should be to keep it unchanged. The burden of proof is on changing it. The design should accommodate the URL, not the other way around.
Where URL changes are unavoidable, the redirect mapping process needs to be treated as a first-class deliverable: a comprehensive list of old URLs mapped to new URLs, implemented before launch, verified in Search Console after launch. This is not a task that can be delegated to post-launch cleanup. Every day a redirected URL spends as a 404 before the redirect is in place is a day that Google is recording the page as gone.
Content decisions that sacrifice rankings for aesthetics
The most consequential redesign decisions for SEO are often not technical. They are editorial. A page that ranks because it contains 2,000 words of detailed, useful content on a specific topic will lose that ranking if the redesign replaces it with 400 words of polished copy and a nicer layout. Google does not reward visual improvement. It rewards content depth, relevance and user engagement signals.
This creates a tension in most redesign projects. The design team is trying to create pages that look clean and communicate quickly. Content that is deep enough to rank tends to look like it would benefit from being shorter. Both things can be true. The resolution is page architecture that accommodates content depth without compromising visual quality: expandable sections, tabbed content, well-designed long-form layouts that read well without feeling like walls of text.
The alternative, which is to simply cut the content because it looks better that way, produces pages that are aesthetically superior and rank for nothing within three months of launch. The business that had a page ranking for a valuable category keyword now has a better-looking page that gets no organic traffic. The design improved. The commercial outcome did not.
The speed regression risk
New design tends to mean heavier pages. More assets, more scripts, more animation, more visual complexity. The old site, for all its visual limitations, may have been loading in 1.8 seconds. The new site, with its improved design, animated sections and additional third-party scripts, loads in 3.4 seconds on a standard NZ broadband connection and 5.2 seconds on mobile.
Google's Core Web Vitals are a documented ranking signal. A redesign that degrades page speed degrades rankings. For NZ businesses competing in specific search categories, a meaningful speed regression can push a page from the third result to the eighth result on a query where the difference in click-through rate between those positions is roughly 800 percent.
The test is simple. Measure the existing site's Core Web Vitals score before the redesign begins. Treat that score as a minimum acceptable baseline for the new site. If the new design cannot meet the baseline without performance engineering work, that work is part of the scope. A redesign that improves appearance at the cost of search performance is a trade the business did not agree to make and probably would not if the cost were made explicit.
What to watch after launch
Even a well-executed redesign creates a period of volatility while Google processes the new site. The question is how to distinguish normal volatility from actual ranking damage that requires intervention.
The signal to watch is not total traffic, which can be noisy. It is the ranking positions of the specific pages identified in the pre-redesign audit. If a page that was ranking third for a target keyword drops to fifteenth within four weeks of launch, that is not normal volatility. It is a signal that something about the page changed in a way Google did not respond well to, and it is worth investigating before the ranking erodes further.
Common post-launch causes include redirects that were implemented but are responding too slowly, content that changed in the migration in ways that affected relevance, internal linking changes that reduced the number of pages pointing at a key page, and indexing delays caused by a sitemap that was not resubmitted after launch. All of these are diagnosable and correctable within a short window after launch. They become harder to correct the longer they persist.
The post-launch period for any significant redesign should include a structured monitoring protocol: weekly ranking checks on the priority pages identified in the pre-redesign audit for at least three months, Search Console crawl reports reviewed for error spikes, and a defined escalation point where investigation begins rather than hoping for recovery.
The redesigns that hold their rankings
The pattern across well-executed redesigns is consistent. The ranking audit happens before design begins. URL changes are minimised and every change that is made is mapped and implemented before launch. Content depth on ranking pages is preserved or improved. Page speed is benchmarked against the existing site and treated as a constraint on design decisions. Post-launch monitoring is structured around specific pages rather than aggregate traffic.
None of this is technically difficult. It is all discipline and prioritisation. The agencies that produce redesigns where rankings hold have made these things part of their standard process. The agencies that produce redesigns followed by traffic drops have not. For the NZ businesses on the receiving end of the outcome, the difference is not aesthetic. It is commercial.
A website redesign is a legitimate risk to an asset the business has invested years in building. That risk is manageable. Managing it requires treating SEO as a constraint on the redesign process rather than a task to address after the new site looks right.
