How to Measure Whether a Website Redesign Actually Grew Your Business

Most redesigns get judged on vibes. The site looks better, the owner is happy, and nobody ever checks whether it made more money. That is a mistake, because a redesign is an expensive bet and you should know if it paid off. Here is how I actually measure one.

Start with the number you already care about

Before touching a single page, write down one primary metric. Not five. One. It should be the thing that, if it moved 20%, would make the project obviously worth it. For most small businesses that is one of these:

• Qualified leads per month (form fills, calls, booked consults)
• Revenue from organic search
• Checkout conversion rate
• Average order value

Everything else is a supporting metric. Bounce rate, time on page, and pages per session are diagnostic tools, not goals. A visitor who finds your phone number in four seconds and calls you has a terrible “time on page” and is your best customer of the day.

Set up tracking before launch, not after

This is where most projects fall apart. If you launch the new site and then realize you cannot compare it to the old one, you have already lost the data. Two things need to be true before the switch.

1. Both versions report to the same property

If you are on GA4, keep the same measurement ID across the migration so history stays in one place. If the old site used Universal Analytics or a different tool, you are starting fresh, so say so out loud and set the baseline date as launch day.

2. Conversions are defined as events, not pageviews

A “thank you” page view is fragile. People hit refresh, they bookmark it, they land on it from a stale link. Fire an explicit event instead. In GA4 that looks like this:

// Fire once, on successful form submission
gtag('event', 'generate_lead', {
 form_id: 'contact-main',
 value: 1,
 currency: 'USD'
});

For phone calls, use a call tracking number or at minimum a tel: click event, since mobile tap-to-call is invisible to analytics otherwise:

document.querySelectorAll('a[href^="tel:"]').forEach(function (link) {
 link.addEventListener('click', function () {
 gtag('event', 'click_to_call', { link_url: link.href });
 });
});

Give it a real measurement window

Do not judge a redesign in week one. You need at least a full month, and honestly 60 to 90 days, before the numbers mean anything. Reasons:

• Google needs time to recrawl and re-rank new URLs
• Returning visitors have to unlearn the old layout
• Seasonality will wreck a short comparison (comparing December to January tells you nothing)

The cleanest comparison is year over year for the same calendar months, not month over month.

Watch the leading indicators while you wait

Revenue lags. These move faster and tell you if you are on the right track:

• Core Web Vitals in Search Console (LCP, INP, CLS). INP replaced FID in 2024, so if your reporting still says FID, update it.
• Form start rate versus form completion rate. A big gap means the form is intimidating or broken on mobile.
• Scroll depth on your main service pages.
• Exit pages, which usually point straight at the weak spot.

Do not skip the redirect audit

If URLs changed, every old URL needs a 301 to its closest equivalent. Not to the homepage. A homepage redirect tells Google the page is gone, and you lose whatever ranking it had. Check for redirect chains and loops too, since they slow the crawl and waste link equity:

# Quick check for a redirect chain
curl -sIL https://example.com/old-page | grep -iE '^(HTTP|location)'

If you see more than one HTTP/301 before the 200, collapse it into a single hop.

Read the results honestly

Sometimes the redesign wins on conversion but loses on traffic for a few months because of URL churn. Sometimes it wins on traffic and loses on leads because the new design buried the call to action. Both are fixable. What is not fixable is not knowing, which is what happens when nobody set up the measurement.

If you are planning a rebuild and want the tracking done right the first time, that is a big part of what I do with Long Island web design work. The build is the easy part. Making it provable is the part that matters.

FAQ

How long before I can tell if a redesign worked?

Give it 60 to 90 days and compare year over year for the same months. Anything shorter is noise, especially if URLs changed and Google is still recrawling.

Is a drop in traffic right after launch normal?

Yes, a temporary dip is common when URLs change, even with good redirects. Watch Search Console for coverage errors and confirm your 301s are single-hop. If it is still down after 8 to 10 weeks, dig into the redirects and internal links.

What if I never tracked conversions on the old site?

Then treat launch day as your new baseline and be honest that there is no before-and-after. Set up events for form submits and calls now, and start the clock. You cannot retroactively create data you never collected.