The urge to tear everything down and do it properly is understandable, but not always justified. A new project means design, development and content all over again, plus a risk to the traffic you've built up. Upgrading a website often achieves the same result faster if the project has a sound foundation. To decide on facts rather than emotion, weigh four factors:
Rebuilding becomes cheaper when the site runs on an outdated platform and the list of required changes touches most of its pages. Otherwise, a staged upgrade makes more sense: the live version keeps bringing in leads while improvements roll out step by step.
Not all improvements are equally useful. Some pay for themselves first because they stop leads from leaking away:
These fixes don't require rebuilding the site, and their effect shows up in analytics within the first few weeks. That's why we put them at the top of the improvement plan and schedule larger features, such as a customer account area or a new integration, afterwards.
Search engines get used to a site: its URLs, headings and page content. When modernising a website, it's important not to destroy what Yandex and Google already rank it for. Before work begins, export a list of all pages that receive traffic, with their titles, H1s and approximate text volume. This is the baseline you'll compare results against.
If moving to new URLs or a new CMS can't be avoided, prepare for it in advance. Build a map matching old and new URLs and set up page-by-page 301 redirects from it, rather than sending everything to the homepage. Keep the staging version closed from indexing to avoid duplicates. After launch, update the sitemap, check Yandex Webmaster and Google Search Console, and keep tracking crawl errors and rankings for key queries for several weeks. Small fluctuations after a migration are normal; a sharp drop is a sign that a step was missed.