When a redesign is the wrong answer

Owners almost always describe the same symptom. Business is slower than it should be, the website looks dated, so the website must be the problem. It is the most available explanation and it is usually the wrong one.

The reason it persists is that a redesign feels like action. You get meetings, mockups, and a launch day. Fixing the actual bottleneck often looks like an afternoon of unglamorous work with nothing to show a partner.

A website is a piece of equipment. It gets judged on output, which is a call, a booking, or a form. The question is never whether it looks current. The question is what it converts, and whether anybody is arriving to convert.

How do I tell whether the site is the problem?

Look at two numbers before you look at a single mockup.

How many people visited last month, and how many contacted you. If a hundred people visited and two called, you have a conversion problem and a redesign might genuinely help. If eleven people visited, you have a traffic problem and the most beautiful site in your county will produce roughly eleven visitors' worth of business.

That is the whole diagnostic. Everything else is decoration on top of it.

The trap is that most small businesses cannot answer either question, because tracking was never configured or broke during the last rebuild. If you do not know your current numbers, the first project is measurement, not design. Rebuilding without a baseline means you will never know whether the new site was better or worse. You will only know whether you like it more.

There is a second reason the website is often the wrong suspect. Whitespark's Local Search Ranking Factors report, published 6 November 2025, had 47 local search experts score 187 factors. The top four for the local pack are primary Google Business Profile category at 227 points, proximity of your address to the searcher at 225, keywords in the profile's business title at 223, and a physical address in the city being searched at 213. Not one of those is your website. The highest website item is keywords in the landing page title tag, 17th, at 146. That is expert opinion rather than a controlled test, and Whitespark says so, but the direction is clear: if you are invisible in the map results, a rebuild is aimed at the wrong asset.

What does the arithmetic actually look like?

What should I actually expect a redesign to move?

Less than the pitch, and this is the part worth being blunt about.

The best published evidence on how much a website change moves a metric comes from Kohavi, Deng, Longbotham and Xu, in a peer reviewed KDD 2014 paper generalising from thousands of controlled experiments at Microsoft, Amazon, Booking.com and LinkedIn. Their finding on effect size, verbatim: at Bing, "most fail, and those that succeed improve key metrics by 0.1% to 1.0%, once diluted to overall impact." They also report that roughly one in 500 experiments meets the bar of a large, replicable, high return change, and that the success rate on ideas generally runs about 10 to 20%.

Then they show what that does to a claimed result. With a one in three prior success rate, a statistically significant finding has an 89% chance of being a true positive. With a one in 500 prior, that same finding has a 3.1% chance. A big claimed lift from a low prior is not evidence, it is noise wearing a percentage sign.

Say the domain gap out loud. That is desktop data from search engines and large retailers, not from a plumber's website, and the authors state their own limits plainly. What transfers is the shape of the world: real changes are small, and a vendor promising a new site will transform your inquiry count is quoting a number nobody measured. Ask what they will measure, over what window, against what baseline.

What problems get misdiagnosed as a design problem?

Four, over and over.

Nobody can find you. Your site is fine, you are simply not appearing in the searches that matter. Sometimes the reason is uncomfortable: businesses that do not exist are ranking above you. Fake listings in the map pack and what to do about them is a real and common cause of a ranking slide that has nothing to do with your website.

The site loads slowly or forms fail silently. This looks like a design problem because visitors leave. It is a maintenance problem, and it is often a hosting problem. Hosting decisions for a small business site get made once, badly, and then quietly cost you for years.

The site is missing pages, not missing style. If you sell six services and have one page that lists all six, nothing you do visually will make you rank for any of them. That is a structure gap, and how many pages a small business site actually needs is an easier question to answer than most owners expect.

The brand is unclear. If you cannot say in one sentence who you serve and what you do differently, a designer will make something attractive that says nothing. That is not a design failure, it is a brief failure, and what you need to settle before a website exists has to happen first or you will pay twice.

What about the speed pitch?

Speed is real and the numbers used to sell it are usually mishandled.

Google's own mobile benchmarks page, the live edition updated in February 2018, reports that Google Research sampled 11 million mobile web domains on a 4G connection in January 2018 and found 79% of pages over 1MB, with 25% able to save more than 250KB by compressing images and text alone. That is the finding that decides your project: most speed problems are page weight, not the platform the page sits on. You do not rebuild a house to move a heavy sofa.

Grade the result against Google's published thresholds rather than a score out of 100. Largest Contentful Paint is good at 2.5 seconds or below, Interaction to Next Paint at 200 milliseconds or below, Cumulative Layout Shift at 0.1 or below, each at the 75th percentile of visits, and all three must pass. If your current site already passes, speed is not your reason to rebuild.

Is there a version of this that is worth doing?

Yes, and it is much smaller than a redesign.

Most sites that feel broken have four or five specific faults. A phone number that is not tappable on a phone. A contact form with nine fields. Page titles that all say the business name. Tracking that never fired. Photos so heavy the page takes six seconds on cellular.

Fixing those is a repair job on a house you already own. It takes days, not months, and it does not put your rankings at risk the way a full rebuild does. A redesign resets everything, including the things that were working.

There is one particular repair that pays out of proportion to its size. Most businesses send every converted visitor to a blank confirmation screen, which is the single most attentive moment a customer will ever give you. Everything a thank you page is wasting is a twenty minute fix that survives whatever you do to the rest of the site.

Do the repairs. Measure for sixty days. Then decide whether you still want the rebuild, because a meaningful number of owners find they do not.

What if the site really is the problem?

Then rebuild, but rebuild for a reason you can state in a sentence.

Good reasons: the business changed and the site describes the old one. You need thirty pages and the platform makes that impossible. The brand is genuinely embarrassing in front of the customers you now want. You do not control the files and the current owner will not release them.

Bad reasons: a competitor's site looks nicer. It has been three years. A salesperson said your site was not mobile friendly without opening it on a phone.

When the reason is real, do the whole job. A rebuild that keeps the old copy, the old photos, and the old structure is a paint job with a launch party.

Where does this advice break down?

Three places, and the third causes the most arguments.

When the site is actively broken rather than merely dated. A form that never delivered, a site that fails on a phone, a platform nobody can log into. Fix or replace it, and do not run a sixty day observation period on a machine that is switched off.

When the ceiling is genuinely being hit. If you are converting well and traffic is real, further site work has little room left. The next dollar belongs somewhere else, usually in the profile, the reviews, or capacity.

When the drop was not the site's fault. Search updates ship constantly, seasons move, a competitor opens. If your traffic fell in a month when you changed nothing, look outside the site first. Rebuilding in response to an external move is how businesses spend a year's budget on weather.

What to do this week

Pull your traffic and inquiry numbers for the last three months. If you cannot, install analytics today and start the clock.

Run the five step calculation above and write down your ceiling. Take that number into any conversation with a vendor.

Open your own site on your own phone, on cellular. Try to call yourself. Try to submit the form. Time the homepage. Write down every failure.

Search the term you most want to rank for, in an incognito window, from your actual neighborhood. Note who is above you and whether they are real businesses.

Then look at your list. If it is full of repair items, you have a repair project. If it is full of "we do not have a page for that," you have a build project. Conflating them is how owners spend the build money on the repair problem.

Questions about whether you should rebuild?

Email me at eric@seod.com with your website address and the three things that bother you most about it. I will sort them into repairs and rebuild reasons and send the list back, plus which two I would do first. If the honest answer is that your site is fine and something else is costing you, that is what I will say.

More on diagnosing this in the web design and branding library.