Many website redesigns don't factor in the decisions that should be led by analytics and business metrics — which pages earn leads, which rankings are worth money, what the site converts at today. That's how a redesign quietly destroys years of accumulated value. Most of the damage is preventable in an afternoon; here's what to get right first.

1. Know what you're protecting

Most redesign plans list what needs to change. Almost none list what must not break. That's backwards — protecting existing value is the cheapest insurance in the project.

Content that earns leads. Find every page that has generated an enquiry, a call, or a form submission. Judge by conversions, not traffic: a page with 50 visitors and 3 enquiries is worth more than one with 5,000 visitors and none. Gets organic traffic and has produced a lead? Preserve list. Everything else — thin content, old SEO plays, retired services — gets flagged for removal.

Rankings worth money. Document your top 20 pages by organic traffic over the past 12 months. Separate real visibility from vanity: ranking for your own business name means little; ranking for "commercial fit-out contractor Brisbane" or "family lawyer Parramatta" is worth real money. And if a page ranks well but converts poorly, fix the page — don't kill it. Replacing a ranking page with a new URL and no redirect is one of the most expensive mistakes in any redesign.

Every lead path, by name. Form submissions, call tracking, Google Business Profile clicks, referral links — map them as named, individual paths, not "our contact form". Redesigns have broken every one of them before, and specificity is what makes post-launch testing catch it.

2. Record your baseline

You can't tell whether a redesign worked if you don't know where you started. Four numbers, documented before anything moves:

  • Organic traffic by page — trailing 90 days, top 50 pages.
  • Form submissions by page — monthly average, per page.
  • Rankings for your top 10–15 non-branded terms.
  • Conversion rate on your primary enquiry action. Converting at 2% today? That's your target to beat.

A redesign you can't measure isn't a redesign. It's a guess with a new coat of paint.

3. Make the technical decisions deliberately

Map every redirect. Every URL with organic traffic that changes address needs a 301 to the correct new page — that's how search engines transfer ranking credit. Skip it and the new pages start from zero. Domain changes raise the stakes further: without a complete map, expect a real traffic drop and months of recovery. Write the map — old URL to new URL — before anything is built.

Change URLs on purpose, not by accident. Cleaner URLs are a fine goal, but each one is a deliberate change with a matching redirect — not a side effect of a template rebuild. "We want cleaner URLs" isn't an instruction. "/services/commercial-plumbing/ becomes /commercial-plumbing/, here's the redirect" is.

Protect your analytics through the switch. A classic failure: the tracking code sits in a template element that gets removed in the rebuild, and half the site silently stops reporting for weeks. Staging hides this — nobody watches staging data. Verify tracking fires on every template before the new site goes live, not after.

4. Fix scope before you start

Redesigns run over when scope gets decided mid-build. Sort every candidate item into three buckets first:

  • Must fix now — actively costing leads or credibility today. A broken form on your highest-enquiry page.
  • Should fix this phase — important, but won't move lead generation on its own. Updated photography, refreshed copy.
  • Can wait — nice to have, or dependent on work that isn't ready. A feature nobody in your market asked for.

Then one filter: has this been tested in your business, or is it untested? A form that converts but looks dated beats a shiny feature nobody's tried. Untested ideas belong in phase two.

Platform and CMS changes are the most expensive scope decisions of all. Before switching, answer one question precisely: what business outcome is the current platform failing to deliver? No precise answer — keep the platform and spend the budget on what's actually broken. A genuine reason to switch — decide it up front. Platform changes made mid-build are how budgets blow out.

5. Know what stops a launch cold

The most common cause of a post-redesign lead drop isn't the redesign — it's what got missed at launch. Each of these takes minutes to check and is expensive to skip:

  • Every redirect works on the live site — correct page, single hop, canonical URL. Chains leak ranking authority.
  • Every form and lead path, tested end to end on the live domain. Staging misses things: reCAPTCHA off in production, a CRM that hasn't whitelisted the new server, forms still posting to a staging endpoint.
  • Staging was never indexed. Check Search Console and remove any staging URLs, or the new site competes against a copy of itself.
  • Analytics fires on every template on the live domain. Don't let the first week of missing data break the news.

If any check fails, the launch waits. None are hard to fix. All are expensive to ignore.

If you'd like a senior view on your redesign before it starts — what's worth protecting, what actually needs to move first, where your scope is likely to creep — get in touch. A focused conversation, not a pitch.