Replatforming a large B2B website without losing the pipeline
A website rebuild for a large ERP brand, run as a revenue system: protect the SEO equity, keep the lead paths, and tie the new site to pipeline instead of redesigning for its own sake.
GTM Architect & Growth Operator · Now · 28 June 2026
TL;DR · Key insights
- A large legacy site holds years of SEO equity and lead routing. A rebuild can wipe both in a weekend
- Score the project on surviving pipeline, not on how the homepage looks at launch
- Measurement, a ranked redirect map, and tested lead paths come before a single new pixel
- Migrate pages by the pipeline they earn, not by how many there are
A large B2B brand decides its website is old and rebuilds it. The project gets scoped as a design exercise: new look, new CMS, launch date. That framing is the mistake.
A mature site for an ERP or business-software brand is not a brochure. It is years of compounding SEO equity, hundreds of URLs that quietly rank, and lead paths wired into a CRM. A rebuild can erase all of that in a single cutover. The redesign looks better on Monday and the pipeline is down by Friday.
I run these rebuilds as a revenue system, not a design project. The goal is simple to state and hard to hold: ship a better site without losing a single point of pipeline, then grow from there.
What the website actually is
Most stakeholders picture the site as a catalogue of pages. That picture is why rebuilds go wrong. The useful picture is a surface that turns attention into a traceable next step.
How the site is treated
Pages owned by nobody. Success measured in launch screenshots. SEO and forms treated as someone else's problem after go-live.
Scored on how it looks
Every template tied to a measurable next step. Rankings and lead routing protected as first-class deliverables, tested before cutover.
Scored on pipeline that survives
The system, in the order it has to happen
The sequence matters more than any single task. Doing the right things in the wrong order is how teams lose rankings they did not need to lose.
-
Measurement before pixels. Before anything is designed, the analytics and CRM have to tell the truth: clean events, a working GA4 setup, conversions that map to real pipeline stages. If you cannot see pipeline before the rebuild, you cannot prove you kept it after.
-
The ranked redirect map. Crawl the entire old site. Rank every URL by organic traffic and assisted conversions. The pages that earn get a 1:1 redirect to their new equivalent. This single artefact protects most of the SEO equity, and it is the step most rebuilds skip.
-
UX and CRO on the money templates. A handful of templates carry the pipeline: the high-intent product and solution pages, the comparison pages, the demo and contact paths. Those get the real conversion work. The long tail does not need a redesign, it needs to keep ranking.
-
Lead paths that survive cutover. Forms, routing rules, tracking, and CRM handoff are rebuilt and tested on staging with real submissions before launch, not discovered broken in production.
-
Content migration by value, not by volume. Migrate the pages that earn. Consolidate the thin ones into stronger pages. Retire what has not been touched or visited in years. Fewer, stronger pages beat a faithful copy of the old mess.
-
An integrated launch, not a big bang. The site goes live alongside the measurement, the redirects, and the campaigns that point at the new pages, so the rebuild lands as one coordinated motion instead of a risky midnight switch.
- Crawl and rank the old site before designing anything
- Redirect every page that earns traffic or assists conversions, 1:1
- Test forms, routing, and tracking on staging with real submissions
- Migrate by pipeline value and consolidate the thin pages
- Keep a before-and-after dashboard the whole team can read
- Treat SEO and forms as a post-launch cleanup task
- Carry every legacy URL across out of caution
- Let the redesign's look decide the information architecture
- Cut over without a rollback plan
- Measure success in screenshots
What to watch after cutover
The first two weeks after launch decide whether the rebuild was a win or a slow leak. The dashboard is small on purpose.
A rebuild that holds these numbers flat through cutover and bends them upward over the next quarter is a success, even if nobody outside the team notices the new design. That is the point. The website that wins is boring: it protects what already earns, then improves the few pages that actually move pipeline.
If your site is due for a rebuild and the plan is mostly about how it will look, that is the moment to stop and re-scope it around the pipeline first. The design is the easy part.
