Setting KPIs for a Website Redesign Project

A company spends three months and twenty thousand dollars redesigning their website. It launches. The stakeholders gather to assess whether it was successful.

Some say it looks modern and impressive. Others say the old site was fine. Nobody can agree on whether it actually worked because nobody defined what "working" meant before they started.

This is almost standard. Most website redesigns happen without clear success criteria. People disagree about metrics, budget, timeline, and what actually matters. The redesign gets done, launches, and then everyone argues about whether it was worth it.

You can avoid this entirely by defining KPIs—Key Performance Indicators—before you redesign. Not vague goals like "increase engagement." Specific, measurable targets that answer: "How will we know this redesign was successful?"

Why you need KPIs before, not after

Without KPIs, a redesign is subjective. "Do you like it?" Some will. Some won't. "Is it better?" Than what? Better at what?

The answer to those questions depends on your business. For an e-commerce site, success might be increasing average order value or reducing cart abandonment. For a software company, success might be getting more demo requests. For a service business, success might be increasing phone calls or email inquiries.

A beautiful redesign that doesn't move the needle on what actually matters is a waste of money. A less aesthetically impressive redesign that increases conversions by 30% is a success, even if some stakeholders prefer the old design.

KPIs force you to be specific about what matters, and they give you something objective to measure against later.

What metrics actually matter for a redesign

There are dozens of metrics you could measure. The goal isn't to track all of them. The goal is to pick the 3–5 metrics that directly connect to business outcomes.

Conversion rate: The percentage of visitors who do the thing you want them to do. For e-commerce, that's completing a purchase. For a services business, that's filling out a contact form or calling a phone number. For SaaS, that's signing up for a free trial. This is usually the most important metric because it directly impacts revenue.

Time to conversion: How long does it take from landing on the site to completing the target action? A slow redesign that drives conversions in 10 minutes might be worse than a fast one that drives them in 30 seconds.

Page load time: Speed directly affects conversion rate, SEO rankings, and user experience. If your redesign is slower than the old site, you've probably made things worse overall even if it looks better.

Bounce rate: The percentage of visitors who leave without taking any action. A lower bounce rate usually means better engagement, though what "good" is depends on your industry and page type. A homepage might have a 40% bounce rate that's fine. A product page might have 70% that's concerning.

Average session duration: How long does a visitor stay on your site? Longer usually means more engaged, but it depends on your goals. If you want quick conversions, a long session might mean people are browsing but not buying.

Return visitor rate: What percentage of your traffic is people coming back for a second visit? Returning visitors usually convert better, so increasing this is valuable.

Page-specific metrics: Depending on what you're redesigning, specific pages might have their own targets. New customer onboarding flow? Track completion rate. Product pages? Track product view depth (how many products do people view before leaving).

User satisfaction: This is subjective, but you can measure it. Post-launch surveys asking "Did you find what you were looking for?" or "How would you rate your experience?" give you qualitative feedback. Aim for a 7+ rating on a 10-point scale.

Pick the 3–5 that matter most to your business. For most companies, that's conversion rate, load time, and bounce rate.

How to set realistic targets

Now that you know what you're measuring, you need targets. Where should these metrics be after the redesign?

This is where most companies go wrong. They set ambitious targets that are driven by wishful thinking, not data. "We want to increase conversions by 50%." Why? What's realistic?

The answer starts with your baseline.

Measure the current site: Use Google Analytics to find your current conversion rate, bounce rate, average session duration, and average page load time. Document these numbers clearly. This is your starting point.

For a site you haven't measured before, collect 2–4 weeks of data before starting the redesign. You need a baseline to compare against later.

Set realistic improvement targets: A well-designed redesign might improve conversion rate by 10–20%. Significant improvements (30%+) are possible but rare and usually require not just a design change but also testing and refinement over time.

If your conversion rate is currently 2% and you want to get to 5%, that's a 150% improvement. That's possible but would usually require more than just a redesign—it might require changing your offer, adding social proof, restructuring your forms, or redesigning multiple pages as an interconnected funnel.

Set targets that feel ambitious but achievable. For most sites, small improvements are wins.

Benchmark against industry standards: If you don't have much data to work with, look at industry benchmarks. E-commerce sites typically convert 1–3% of visitors. B2B software leads might convert 5–10% of visitors to free trials. Services businesses might get 2–8% of visitors to fill out a contact form.

Don't use these as your targets, but use them as context. If you're at 0.5% when the industry average is 2%, you have room to improve. If you're at 4% when the average is 2%, you're doing better than average—set conservative targets.

Separating design improvements from everything else

Here's an important detail: A website redesign affects metrics, but so do other factors. Your search engine rankings change. Your paid ads audience shifts. Your email marketing campaign runs. The season changes. Your competitor launches a feature.

If you measure only the pre-launch and post-launch numbers, you won't know whether the redesign actually caused the changes.

To isolate the impact of the redesign:

Measure before and after against a time period, not a date: Compare the 30 days before launch to the 30 days after launch. This controls for daily fluctuation.

Segment your traffic by source: Direct traffic, organic search, paid ads, social, etc. Changes in paid ad performance don't indicate the redesign worked. Changes in direct traffic (people who know where you are) and organic search (long-term SEO impact) are better indicators.

Run an A/B test if possible: For critical pages, consider keeping the old design for 10–20% of traffic while rolling out the new design to 80%. This lets you compare performance side-by-side. After you have confident data, switch everyone over.

Note external factors: If you launched during a holiday season, or your competitor launched a feature, or you ran a big marketing campaign, document it. These are confounding factors that might explain changes more than the redesign does.

When and how to measure

Before launch: Take your baseline numbers. Document conversion rate, bounce rate, load time, whatever you're measuring. Take screenshots. Get them in writing. You'll want to refer back to these.

After launch - first week: Check your metrics daily. Look for bugs, crashes, or unexpected drops in traffic. The first week often reveals problems that need immediate fixing.

After launch - first month: Week-by-week comparison. Are your target metrics improving, staying flat, or declining? If you're going the wrong direction, figure out why. Maybe the redesign introduced usability issues. Maybe you need to update your calls-to-action. Don't wait three months to find out.

After launch - ongoing: Continue measuring. Most redesigns see an initial bump (people are curious about the new site), then settle into a new baseline. Give yourself 60–90 days to see the real impact.

When to redesign again

This is important: If your redesign hits its targets, great. Don't immediately redesign again. Use your time and budget to test variations, optimize what's working, and collect feedback from real users.

If your redesign misses targets significantly—say you aimed for 3% conversion and you're at 1.8%—investigate why before redesigning again. Is it a design problem or a messaging problem? Did you measure correctly? Are there confounding factors you missed?

A redesign is expensive. Measuring the impact and learning from it is how you eventually get better at it.

Getting buy-in on KPIs before you start

Setting KPIs requires conversation. Different stakeholders care about different things. Your CEO cares about revenue impact. Your marketing team cares about traffic. Your design team cares about aesthetics. Your engineering team cares about load time.

These aren't conflicting—they all matter. But you need agreement on which ones matter most.

Bring the stakeholders together early and ask: "What does success look like for this redesign?" Listen to each perspective. Then ask: "If we had to pick the three most important things, what would they be?"

Get agreement in writing. This becomes your project charter. If scope creeps or priorities shift later, you have the original agreement to refer back to.

The redesign that nobody can evaluate

At the other extreme: A company redesigns their site but doesn't track any metrics. They can't tell you if it worked. They just know they spent money and time. A year later, somebody says "I think the old site performed better" but there's no data to support it.

This is avoidable. Measure now. You'll be able to justify the redesign investment with data. And more importantly, you'll know what to do next.

---

FAQ

Can I set KPIs after the redesign launches? Technically yes, but you lose the ability to compare fairly. You need a baseline from before the redesign. If you haven't measured yet, measure immediately and use those numbers as your new baseline, but you'll have lost the first month of post-launch data.

What if we redesign for reasons other than increasing conversions? Set KPIs that reflect your actual goals. Redesigning for brand refresh? Measure brand perception and aesthetic preference. Redesigning for accessibility? Measure accessibility score and user task completion. Redesigning for speed? Measure page load time and Core Web Vitals. Your KPIs should match your goals.

Should we set the same KPIs for every page? No. A homepage might have different goals than a product page or a blog post. Set specific KPIs for high-value pages. For low-traffic pages, track them but don't stress if the numbers are noisy.

What if my conversion rate drops after the redesign? Investigate. Did you change your call-to-action or form? Did you change the flow? Did something break technically? Did you lose search rankings? Once you find the cause, you can fix it. A small drop that recovers within a few weeks is often just users getting used to the new design.

How many KPIs should I track? 3–5 is usually right. Track more than that and you lose focus. Track fewer and you might miss important changes. Most companies track conversion rate, load time, and bounce rate. Add page-specific metrics if needed.

Should I measure user satisfaction alongside conversion metrics? Yes. Sometimes a design that converts well doesn't feel good to users, and they abandon you later. A simple post-launch survey asking "How would you rate this experience?" takes 10 minutes to set up and gives valuable context alongside your conversion numbers.

What if our baseline data is messy or incomplete? Start over. Measure 2–4 clean weeks before redesigning. Yes, this delays the redesign. It's worth it because you need to know what "before" looks like.

If we see good results after the redesign, should we stop improving? No. Continuous improvement is different from redesign. You've got a working baseline—now use A/B testing and iteration to improve further. Small changes compound over time.

What's a "good" improvement after a redesign? 10–15% improvement in your target metric is solid. 20%+ is great. 30%+ is excellent and usually indicates you made significant changes beyond just aesthetics. Even 5% improvement adds up when scaled across thousands of visitors.