Website redesign
A website redesign that fixes what the site does.
A redesign is not a new coat of paint. If the site looks dated, loads slowly, or turns visitors away, the rebuild has to change the structure, the message, and the speed. I baseline the current site, map the pages and flows that matter, build it in code, and re-measure the result.
Answer in 30 seconds
A website redesign is quoted in the Web Development band, $600–$5,000 depending on page count and content. It starts with a performance and funnel baseline, a page map, and a performance budget, then ships in code with a before/after Lighthouse re-run. Fixed USD quote before code.
From $600 · published price band
Not sure this is the right service? Find your fix in three questions →
Fixed USD quote before code · 30-day warranty on the delivered scope
A baseline, not a mood board
The current Lighthouse mobile run, the pages that actually get traffic, and the flows where enquiries are lost. The audit says what to keep as well as what to cut.
A page map with a job for each page
One primary conversion path, a clear hierarchy, and no page that exists only because the old site had one. The sitemap and the copy plan are agreed before design starts.
Designed and built in code
One person owns the design, the build and the launch. A performance budget is set at the start, so the new site cannot ship heavier than the old one.
Before/after, published
The same Lighthouse mobile profile re-run after launch, plus the funnel events that matter. Both numbers are published, including the ones that are not flattering.
What actually changes
A redesign has four jobs. If a proposal only mentions colours and a new template, it is not addressing the parts that decide whether the site works.
The message
One clear sentence about who the site is for and what happens next. The old copy is audited, not copied forward.
The page map
Every page gets a job in the funnel. Pages that only exist because the old site had them are cut or merged.
The performance budget
A first-load byte and LCP target is set before design starts, so the rebuild cannot ship heavier than the site it replaces.
The measurement plan
The same Lighthouse profile and the same funnel events are recorded before and after, so the result is comparable.
What it costs
The published bands come from the same price list as /services. The call picks the band that fits the scope; it does not invent a new one.
Starter Site · $500
A standard 3-page site for a small business that needs a clear, fast presence without a custom content model. 5–7 days.
Web Development · $600–$5,000
A custom site or rebuild: page map, design in code, content structure, performance budget, analytics and a measured handover. 1–6 weeks.
Web Apps · $2,800–$12,000
If the redesign includes accounts, billing, dashboards or a real product workflow, it is a web app, not a marketing site. About 5 weeks for the core loop.
Before you hire anyone: the four-point audit
Run this on the current site before commissioning a redesign — from me or from anyone. It costs an afternoon, and it is the same baseline any honest proposal should start from.
Speed, measured on mobile
Run PageSpeed Insights on the mobile profile and write down the score, LCP and INP. Without a baseline number, “feels slow” can neither be diagnosed nor proved fixed.
The enquiry path, end to end
Submit your own contact form and count what actually arrives. A form that renders but never delivers, or a phone number that is text instead of tap-to-call, loses enquiries silently while the owner believes the form is working.
The page map, page by page
List every page and the one job it does in the funnel. Pages with no job get cut or merged before a redesign starts — not rebuilt out of habit.
Measurement, before anything moves
Check which analytics exist and which events actually fire. A redesign with no before-baseline cannot prove it improved anything, which is why every rebuild here publishes both runs.
The proof
≈228 → 182.9 KB modern first-load
This portfolio's own rebuild: the root motion provider removed, the hero rewritten without the motion runtime, and per-route budgets enforced. The teardown also explains why the old 266.5 KB figure was inflated by a legacy polyfill.
Read the rebuild teardown →74 → 97 · LCP 7.2s → 2.0s
Small Batch Brine concept storefront: a rebuild that removed a 768 KB phantom preload and re-measured the same mobile profile.
Read the Small Batch Brine breakdown →Self-initiated concept builds, no client affiliation implied. Dated Lighthouse mobile runs, including the ones that score badly.
Related services
Related reading
What a website costs in 2026, by project type
Where the $600–$5,000 band sits against landing pages, stores and apps.
What a $500 website costs by year two
Why the cheap rebuild is often the expensive one.
The portfolio's own rebuild
A measured redesign: root motion removed, budgets enforced, before/after published.
FAQ