Before You Redesign Your Website, Do These Nine Things

Most website redesigns are judged on how the new site looks on the day it launches. Six months later nobody has checked whether the enquiries went up, and in a fair number of cases they went down, because the old site's best-performing page quietly disappeared during the rebuild and nobody noticed.

The design is rarely the reason a rebuild succeeds or fails. The preparation is. A website redesign checklist is worth having because almost everything that goes wrong is decided in the two weeks before anyone opens a design tool.

Nine things, in the order you should do them.

1. Write down the numbers you have now#

You cannot prove a redesign worked without a before. Pull four figures and put them somewhere permanent.

Monthly organic sessions, averaged over the last twelve months so seasonality does not fool you. Total enquiries per month, split by phone, form and any other channel you can see. Your ten most visited pages with their session counts. Load time on a mobile connection, tested on an actual phone rather than a desktop browser.

Take screenshots of the analytics view as well as noting the numbers. Analytics platforms change their interfaces and retention windows, and a screenshot survives that.

If you cannot get these numbers because nothing is properly tracked, fix the tracking first and give it a month. Launching a rebuild with no baseline means you will spend the next year arguing about whether it worked.

2. Find out which pages are actually earning#

Almost every service business site has a handful of pages doing nearly all the work, and a long tail doing nothing. The tail is safe to cut. The handful is not, and people cut it constantly because it looks dated.

Sort your pages by organic entries, then again by conversions. A page that ranks for something valuable and has been sitting there for six years is an asset. Rewriting it from scratch during a redesign, changing its URL, and dropping half its content is how businesses lose rankings they cannot get back.

Mark those pages as protected before anything else happens. Protected means the URL does not change and the substance of the content carries over even if the layout does.

3. Build the redirect map before the design#

This is the step that gets skipped, and it is the one that causes the most expensive damage.

Export a full list of your current URLs. Any site crawler will do it, and so will your sitemap file. Put them in a spreadsheet with a column for the new URL each one will point to after launch. Every single old URL needs a destination, and that destination should be the closest equivalent page rather than the homepage. Redirecting everything to the homepage is the same as deleting the pages as far as your rankings are concerned.

Doing this before the design has a second benefit. The map forces a decision about site structure while it is still cheap to change, which is when structure decisions should be made.

Set the redirects as 301s, permanent ones, so search engines pass the accumulated value across. Then test them after launch by working through the spreadsheet row by row. Automated tools will check a sample. The full list is worth an hour of somebody's time.

4. Decide what the site is for, in one sentence#

A redesign brief that says "modern, clean and professional" gives the designer nothing to aim at and gives you no way to reject anything.

Write the actual job instead. Something like: this site exists to turn people searching for emergency plumbing in west Auckland into phone calls, at any hour. That sentence tells you the phone number belongs above the fold, that the emergency page matters more than the about page, and that the after-hours experience needs designing rather than assuming.

If two people in the business would write different sentences here, resolve that before the project starts.

5. Get the content ready before design starts#

Content is what delays rebuilds. Design gets approved in week three, and then the site sits half-built for two months waiting for service page copy that nobody has time to write.

Write the pages first, in a plain document, so the design can be built around real words. Pages designed around placeholder text always have to be reworked when the actual copy turns out to be twice as long. There is a section-by-section structure with worked example copy in writing a service page that actually gets enquiries, which is the fastest way to get service pages drafted without staring at a blank page.

Photography counts as content. Book the shoot early, and get photos of your actual work and your actual team rather than paying for stock images that every competitor also has access to.

6. Check what your current site is doing wrong before you replace it#

A rebuild that carries the old problems into a nicer layout is a common and expensive outcome. Diagnose first.

Look at where visitors leave, whether the offer is clear within a few seconds, whether contacting you takes more than one action on a phone, and whether anything answers people who arrive outside business hours. Your website gets traffic and no enquiries works through the diagnostic in order.

Write the findings into the brief as requirements. "The contact number must be tappable and visible without scrolling on every page" is a requirement a designer can meet and you can check. "Better user experience" is not.

7. Plan what happens after someone clicks contact#

A redesign is the natural moment to fix what happens after somebody decides to contact you, and it is usually treated as out of scope.

Decide who answers enquiries that arrive at 8pm on a Sunday, and how. Decide whether the site is going to answer common questions itself or send everybody to a form and a wait of unknown length.

These decisions change the build. A site meant to qualify and book work needs different components to one meant to display information, and retrofitting the difference costs more than including it. What makes a website smart sets out the capabilities worth specifying, and qualifying leads on your website covers the sorting layer if enquiry volume is already high enough to be a problem.

8. Agree who can edit the site after launch#

Ask this before you sign anything: when I want to change a price or add a service page in eighteen months, what exactly do I do.

If the answer involves emailing the agency and waiting, price that in for the life of the site. Plenty of good builds work that way and it is a fair arrangement when the ongoing relationship is fair. What you want to avoid is discovering the arrangement afterwards, when a two-line change has become a quoted job.

Also settle ownership of the domain and the hosting account in writing, along with the analytics property. These get tangled during rebuilds and they are unpleasant to untangle later. The cost side of all this is covered in what a business website costs in New Zealand.

9. Set the launch-week checks now#

Write the post-launch checklist while you still have attention on the project, because launch week is chaos and nobody thinks clearly in it.

The noindex one deserves a specific mention. Development sites are usually blocked from search engines while they are being built, and the block occasionally survives the move to production. A site can sit invisible for a month before anyone works out why the phone stopped ringing.

The two items on this website redesign checklist to do this week#

Do step one now, regardless of when the redesign happens. Recording your current traffic and enquiry numbers takes an hour and it is the only step that becomes impossible to do later.

Then do step three. The redirect map is tedious, it is the least interesting part of any rebuild, and it is the difference between a site that carries its search visibility across and one that starts again from zero. Both steps happen before you have chosen a designer, which is precisely the point.