Rebuilding a Plumbing Website Without Losing What Already Works

Nearly every rebuild we’re asked to look at starts the same way. The site
looks dated, somebody quotes a redesign, and the conversation is entirely about
appearance. Nobody asks what the current site has already earned.

A prettier website is not automatically a better website. Rebuilding
should begin by learning what the existing site has already earned.

What “already earned” actually means

Search equity is an unglamorous term for something concrete: the accumulated
visibility a site has built up over years. Indexed URLs that rank. Other sites linking
to specific pages. Content that answers a question well enough that people keep finding
it. Internal relationships that tell search engines how the business fits together.
Familiarity, so that people searching your company name find you.

None of it appears on a design mockup. All of it can be discarded in an
afternoon.

The inventory that comes first

Before anything is designed, this is what needs writing down.

• Every indexed URL. Not the pages someone remembers — the
  actual list.
• Which pages get traffic, and for what. Search Console and
  analytics, both.
• Which pages have links pointing at them from other sites. These
  are the ones where a broken URL costs the most.
• Existing redirects. Sites accumulate them. Chains and loops
  inherited from a previous rebuild are common and quietly damaging.
• Metadata. Titles and descriptions that already work.
• Structured data. What the site currently declares about
  itself.
• Internal linking. Which pages support which.
• Forms and tracking. Where enquiries go, and whether anyone has
  tested it recently.

URLs are the part people get wrong

Google’s site-move documentation is the reference here, and its position is
consistent: if URLs change, every old URL needs a permanent redirect to the closest
equivalent new one. Not the homepage — the equivalent page. Redirecting
everything to the homepage is functionally the same as deleting the content, because
the destination doesn’t answer the query the old page answered.

The corollary matters just as much: if you don’t need to change a
URL, don’t. A redirect is a repair for a problem you created. The
cheapest migration is the one where the addresses stay put.

Content: keep, improve, expand

The instinct on a rebuild is to rewrite everything. It’s usually wrong.

If a page already ranks and brings enquiries, the first question isn’t how to
replace it. It’s what made it valuable enough to rank in the first place. Often
the answer is something the new design would have discarded: a long, plain, genuinely
useful explanation that doesn’t fit the new template’s idea of a section.

Three buckets, decided page by page:

• Keep. Working, still accurate, same URL. Leave it alone.
• Improve. Right subject, thin execution. Same URL, better
  content.
• Expand. One paragraph doing the work of a whole service. Give it
  its own page, and link the original to it.

The technical foundation underneath

A rebuild is the moment to fix things that are hard to change later: hosting that
can handle the traffic, a platform someone can actually maintain, HTTPS configured
properly, a sitemap that reflects reality, robots directives that don’t
accidentally block anything, and page performance measured against Core Web Vitals
rather than assumed.

One specific trap: staging sites are routinely built with search engines discouraged,
and that setting is routinely carried to production. It is worth checking on launch day
and again a week later.

Before, during and after

Before

Full backup with a tested restore. The URL inventory. A written redirect map. Agreed
list of pages to keep untouched.

During

Build on staging, blocked from indexing. Verify the redirect map against the real
inventory rather than the intended one.

After

Check indexing status. Crawl the whole site for broken links and redirect chains.
Confirm the sitemap is correct and submitted. Submit a change of address if the domain
moved. Test every form by actually sending one. Re-check performance. Then watch Search
Console weekly for a month, because that is where a missed redirect shows up first.

The FLOWLINE view

We treat a rebuild as a preservation exercise with a design attached, not the other
way around. Sometimes the right recommendation is that the site doesn’t need
rebuilding at all — it needs four pages it doesn’t have and a fix to the
navigation.

That is a less exciting proposal than a redesign. It is also, fairly often, the one
that produces more calls.