By Mehdi HassanHow do you measure whether a website redesign was successful?
You measure a website redesign by comparing the new site's commercial results against a baseline captured before launch, over comparable periods. Success means the site does its job better: more qualified leads, signups or sales from the same or more traffic, without losing search visibility.
Most teams skip the baseline and end up arguing from impressions: the site feels faster, the sales team likes it. Those opinions matter, but they don't tell you whether the money was well spent. This framework gives you the numbers to capture, the goals to set and a review rhythm. It's part of our complete website redesign guide.
Which baselines should you capture before a redesign launches?
Capture baselines for traffic, conversions and site health for at least the 12 months before launch, so you can compare against the same season. Export them to a sheet; dashboards change when tracking changes, and you want a fixed record.
Baseline | Where to get it | Why it matters |
|---|---|---|
Organic clicks and top queries | Google Search Console | Shows whether search visibility held |
Sessions by channel | Analytics | Separates SEO effects from paid and direct |
Conversion rate by key page | Analytics with conversion events | The core measure of the redesign |
Qualified leads or signups | CRM | Filters out spam and poor-fit leads |
Lead-to-customer rate | CRM | Shows whether lead quality changed |
Core Web Vitals on key templates | Search Console or PageSpeed Insights | Speed affects both rankings and conversion |
Top landing pages and their traffic | Search Console and analytics | The pages you must protect |
Save a copy of the old site's key pages as well. When numbers move later, being able to compare old and new versions saves hours. If you're planning a migration at the same time, our guide to building a redirect map covers the URL baseline.
Which website redesign goals and KPIs should you set?
Set one primary goal that reflects why you redesigned, and two to four guardrail metrics that must not get worse. A single goal keeps decisions clear, and guardrails stop you from winning one number by quietly losing another.
Why you redesigned | Primary KPI | Guardrails |
|---|---|---|
More demo requests | Demo conversion rate on key pages | Organic clicks, lead quality |
Self-serve growth | Signup conversion rate | Activation rate, organic clicks |
Moving upmarket | Qualified enterprise leads | Overall lead volume, sales cycle feedback |
Rebrand | Branded search and direct traffic | Conversion rate, organic clicks |
Team can't edit the site | Pages published per month without dev help | Speed, conversion rate |
Write the goal down before design starts. Our article on signs you've outgrown your website helps if you're not sure which problem the redesign is solving.
What should the 30, 60 and 90-day redesign reviews cover?
Each review answers a different question. Day 30 asks whether the site is healthy, day 60 asks whether results are moving the right way, and day 90 asks whether the redesign worked.
Day 30: health check
Confirm tracking fires correctly, forms reach the CRM, redirects work and no key page lost its rankings. Fix problems; don't judge results yet.
Day 60: direction check
Compare conversion rate by page and organic clicks with the same period in the baseline. Identify the two or three pages that underperform and plan changes.
Day 90: verdict
Compare the primary KPI and guardrails over a full quarter against the same quarter before. Decide what worked, what to iterate and what to roll back.
A 90-day review template you can copy
Section 1, the goal: restate the primary KPI and the guardrails agreed before launch.
Section 2, the numbers: a table with baseline, current value and change for each KPI, using the same date ranges and the same definitions.
Section 3, page view: conversion rate and organic clicks for the top ten pages, old versus new.
Section 4, other factors: campaigns, pricing changes, seasonality, product launches and tracking changes during the period.
Section 5, decisions: what to keep, what to test next, and who owns each action.
Section 6, next review: the date and the questions it will answer.
How do you separate the redesign's effect from everything else?
Separate the redesign's effect by comparing like-for-like periods and logging everything else that changed. A redesign rarely launches in a quiet month, and campaigns, pricing changes and seasonality all move the same numbers.
- Compare the same period last year as well as the period just before launch.
- Split by channel. If paid traffic doubled, site-wide conversion rate will change for reasons that have nothing to do with design.
- Keep a change log of campaigns, pricing, product releases and tracking edits.
- Check tracking parity. If a conversion event was renamed or a form was rebuilt, old and new numbers may not count the same thing.
- Look at pages, not averages. One improved pricing page can be hidden by a weaker blog.
If you want to put a money figure on it, our guide to website redesign ROI turns these KPIs into a return calculation.
Who should own redesign reviews after launch?
The reviews need an owner who can both read the numbers and change the site, otherwise the 60-day findings sit in a document. For a lean team, that's often the hardest part of the whole redesign.
Founders and marketing leads usually own the goal. The work after each review, such as rewriting a page, fixing a form event or testing a new hero, needs design and development time. Our marketing website dashboard covers the metrics to keep watching, and if traffic fell after launch, the recovery playbook for post-redesign traffic drops shows what to check first.
Which redesign success metrics should you ignore?
Ignore metrics that can rise without the business improving. They're easy to report and easy to misread.
- Time on page on its own. Longer can mean more engaged or more confused.
- Bounce rate on a single-purpose page. A visitor who books a demo and leaves did exactly what you wanted.
- Total pageviews. Splitting content across more pages inflates it.
- Raw lead count without quality. A looser form can raise lead volume while sales-qualified leads fall.
These can still be useful as signals for a specific page. Just don't put them in the verdict.
