Noibu blog

My Conversion Rate Dropped After a Site Redesign. How Do I Find What Broke?

Megan Glover

October 20, 2026

Blogs

To find what broke after a redesign, compare conversion before and after launch by page, device, and browser, then watch sessions on the pages that fell furthest. Redesigns usually hurt conversion through new errors, slower pages, or controls shoppers can no longer find. Rank each cause by lost revenue, then address the biggest one first.

A redesign is supposed to lift conversion. When the numbers fall instead, the cause is rarely the new look itself. It is usually something specific that changed with the launch: a script that stopped loading, a template that got heavier, or a button that moved out of the place shoppers expect it. The job is to narrow the drop to a page, a device, and a cause before anyone reverts the design.

TL;DR

Split the drop by page, device, and browser before changing anything.
Look for new errors and slower pages that began on launch day.
Watch sessions on the worst pages to see what shoppers actually hit.
Rank causes by lost revenue, not by how loudly the team complains.

How do I know whether the redesign caused the drop?

Line up conversion rate by day against the launch date. If the decline starts on launch day or within a day of it, the redesign or something deployed with it is the likely cause. If it began earlier or drifted down slowly, look at traffic mix, promotions, and seasonality first. Compare the same days of the week, because weekday and weekend behavior differ.

Then check traffic quality. A new campaign or a change in channel mix can lower conversion without any site fault, so compare conversion for each traffic source separately before blaming the design.

How do I narrow the drop to a page, device, or browser?

A redesign touches templates, so problems cluster. Break conversion into the steps of the journey: product view, add to cart, checkout start, and order. The step with the largest relative drop points to the template at fault. Then split that step by device and browser. A fall on mobile Safari only, for example, points to a layout or script problem, not a pricing or messaging one.

What you see	What it usually means	What to check next
Drop at add to cart on one template	A moved or restyled button, or a variant selector that no longer responds	Click behavior and errors on that template
Drop on one browser or device only	A layout, script, or CSS issue specific to that environment	Sessions and errors filtered to that environment
Drop across every step	Slower pages or a site-wide script failure	Core Web Vitals and errors since launch day
Drop only at checkout	A changed checkout flow or a payment integration that broke	Checkout errors and form-field behavior

What should I look for in errors and page speed after a redesign?

Compare the error list from the week before launch to the week after. Errors that did not exist before and appear on launch day are prime suspects, especially on templates that were rebuilt. Next, compare page speed on the redesigned templates. New hero images, extra scripts, and heavier themes often make pages slower even when they look better.

Noibu research found that over 90% of website errors are never reported, so waiting for customers or support tickets to flag a problem after launch means most of them go unseen.

How do I see what shoppers actually experienced?

Numbers tell you where; sessions tell you why. Pull a handful of sessions that reached the weak step and did not convert, and watch what happened: a button that did nothing, a field that rejected valid input, or a page that never finished loading. Three or four sessions that show the same behavior usually confirm the cause.

How do I decide what to address first?

Estimate the revenue at stake for each cause: the number of affected sessions, multiplied by your conversion rate and average order value. A bug that blocks add to cart on mobile outranks a cosmetic mismatch, even if the cosmetic issue is easier to spot. Work down the list in that order and re-measure after each change, so you can tell which change moved the number.

How do Noibu's agents fit in?

Two of Noibu's six supervised agents map directly to this problem. The Bug Resolution Agent identifies the top issues, traces root causes to your codebase, and creates a pull request for you to review. The CRO Agent finds underperforming product pages and rewrites them to convert better, and you review the changes before anything is published. Noibu's agents work in the Noibu console and in AI tools such as Claude Cowork, Microsoft Copilot, and OpenAI Codex, and every agent follows the same loop: find, build, your approval, ship, measure. Nothing ships until a person says so. For a broader look at how the agents split the work, read AI agents for ecommerce: what they actually do in 2026.

Noibu is AI for ecommerce operations, the always-on optimization system that keeps online stores running at their best. A team of supervised AI agents finds what's broken, slow, or leaking revenue, does the work, and hands results back for human approval. Trusted by retailers like Free People, Guess, Liquid I.V., and Ariat.

Frequently asked questions

Why did my conversion rate drop after a redesign?

Redesigns change templates, scripts, and page weight all at once, so a drop usually traces to a new error, a slower page, or a control shoppers can no longer find, rather than to the visual style itself.

How long should I wait before reacting to a conversion drop?

Compare at least one full weekly cycle against the same days before launch. A drop that starts on launch day and persists for several days deserves investigation without waiting longer.

Should I roll the redesign back?

Only after you know what broke. A full rollback discards the parts that work. Isolating the page, device, or error behind the drop lets you correct the cause and keep the rest.

Can a redesign slow my site down?

Yes. Larger images, extra scripts, and heavier themes can slow pages even when they look better, and slower pages convert less. Compare page speed on the new templates against the old ones.

What data do I need to find what broke?

Conversion by page, device, and browser; errors and page speed before and after launch; and sessions from shoppers who reached the weak step without converting.