Before/After: What a Small Business Website Redesign Actually Changes

Search "website redesign case study" and most results are agency portfolio marketing: a client name, a percentage lift, no way to check either one. That's not useful when you're deciding whether your own redesign is worth the disruption. This post skips the unverifiable client logos and works from what's actually documented: a published Nielsen Norman Group case study of a real small manufacturer's navigation redesign, NN/g's own multi-decade research on what usability redesigns typically deliver, and Baymard Institute's ongoing benchmark of what "good" checkout design looks like. None of it is Alphalead's own client work — we don't have a named case study to publish yet, and we're not going to invent one. What follows is what independently published research says a redesign actually changes.

Key Takeaways

• In a published Nielsen Norman Group case study, a small industrial-machinery manufacturer's navigation redesign raised its findability score from 4.0 to 7.4 out of 10 — an 85% improvement — after a structured tree-test and card-sort process.¹
• Across 99 case studies collected over five editions of NN/g's UX Metrics and ROI research since 2006, the average KPI improvement from a usability-driven redesign fell from 247% in the 2006-2008 period to 75% by 2020 — not because redesigns stopped working, but because the internet's easy wins are mostly already claimed.²
• Only 2% of the 344 top-grossing e-commerce sites Baymard Institute benchmarks score "good" on checkout usability, and the average site still has 32 unimplemented improvements sitting on the table.³
• 81% of small businesses have redesigned their website at least once, and 90% plan to invest in their website again in the next 12 months, according to Clutch's August 2025 survey of 406 U.S. small business owners.⁴
• The same survey found 45% of small businesses outsource their website build entirely, and only 12% run something fully custom-built — most redesigns happen on off-the-shelf platforms, not from scratch.⁴

In this guide:

• What small businesses actually do when they redesign
• The problems a redesign is usually fixing
• A documented case: Baileigh Industrial's navigation redesign
• What the research says redesigns actually deliver
• How to approach your own redesign
• FAQ

What Small Businesses Actually Do When They Redesign a Website

Before getting into what a redesign changes, it's worth being clear what "redesign" actually means at small business scale — it isn't a from-scratch rebuild for most owners.

In Clutch's August 2025 survey of 406 U.S. small business owners, 81% said they've redesigned their website at least once, and 90% plan to invest in their site again over the next 12 months.⁴ That's a population treating a website as something to keep revising, not a one-time project to finish and forget — consistent with the "living asset" framing Alphalead's branding guide covers for brand identity generally.

The same survey shows most of that work isn't custom development. 45% of small businesses outsource their website build, 37% handle it in-house, and 8% have the owner build it themselves — only 12% describe their site as fully custom-built, with the rest running on platforms like Wix, Squarespace, WordPress, or Shopify.⁴ Practically, that means most redesigns are a re-templating and content overhaul inside an existing platform, not a ground-up rebuild — which matters for scoping one, since the biggest gains tend to come from structure and content decisions, not from switching platforms.

The Problems a Redesign Is Usually Fixing

A useful redesign is almost always solving one of three problems, and they're worth naming separately because they call for different fixes.

People can't find what they came for. A navigation problem, not a visual one — a site can look polished and still bury what a visitor needs behind a category label that doesn't match how customers actually think about the product. It's also the most measurable of the three, which is why it's the clearest documented example in this post.

People don't trust what they're looking at. Design is the single most-cited factor in how people judge a website's credibility, and that judgment happens before anyone reads a word of copy — covered in more depth in Alphalead's branding guide. A redesign that fixes navigation but leaves a site looking unfinished hasn't solved the trust half of the problem.

The checkout or contact flow loses people who were already convinced. This is the most quietly expensive version, since it costs you visitors who'd already decided to buy. Baymard Institute's ongoing benchmark of 344 top-grossing U.S. and EU e-commerce sites — more than 30,000 checkout elements scored against a research-backed standard — finds only 2% reach a "good" rating, and 0% reach the top tier, with the average site still carrying 32 unimplemented improvements.³ That study covers large e-commerce sites, not small businesses, but the pattern — even well-resourced companies leaving real conversion on the table in a flow they assume is "done" — is the same failure mode a small business contact form runs into, just at smaller scale.

A Documented Case: How Baileigh Industrial Redesigned Its Website Navigation

Most of what's written about redesign results is either a large enterprise case study or an agency's self-reported client win. This one is neither: it's a published Nielsen Norman Group case study of a genuinely small business, with a methodology detailed enough to check.

The situation. Baileigh Industrial sells industrial metalworking and woodworking machinery direct to customers and through distributors on Baileigh.com, with individual orders often running over $15,000 — a specialized manufacturer, not a national retailer dressed up as small-business advice.

The problem. Customers repeatedly struggled to find the product category they wanted and wasted time going down the wrong navigation path before landing on the right page — not a cosmetic issue on a site where one wrong click can cost minutes on a five-figure purchase decision.

The approach. Marketade, the agency handling the redesign, ran a tree test using Optimal Workshop's Treejack with 64 participants recruited through Baileigh's newsletter and social channels, testing 8 tasks built around top-selling categories — things like finding where to bend quarter-inch sheet metal. A card-sort study followed, to learn how customers naturally grouped products, and the site's information architecture was rebuilt around those groupings: broader top-level categories, clearer subcategory names, less reliance on one first click sending someone down the wrong path.

The outcome. After launch, the same 8 tasks were re-tested against a fresh participant sample using the new navigation.

Metric	Before	After	Change
Findability score (avg. across 8 tasks, out of 10)	4.0	7.4	+85%
Tasks scoring 7 or higher (out of 8)	0	6	+6 tasks
Web revenue and leads (Marketade's own reporting, not independently audited)	Baseline	"Up by large margins"	Directional only

Worth being precise about that last row: the revenue and lead claim comes from Marketade, the agency that did the work, quoted in NN/g's published writeup — not an independently audited figure. The findability scores above it are the actual measured result, from a controlled before/after test on the same 8 tasks. That's the honest way to read most redesign "results": treat the measured methodology as the real evidence, and treat self-reported revenue commentary next to it as directional color, not proof.

What the Research Says Redesigns Actually Deliver

One case study is a data point, not a pattern. For the pattern, Nielsen Norman Group has aggregated outcomes across many redesign projects since 2006, through five editions of its UX Metrics and ROI research covering 99 case studies.²

The trend is the most useful part. In the 2006-2008 period, redesigns in NN/g's dataset delivered an average KPI improvement of 247%. By the 2020 edition, that average had fallen to 75%.² NN/g's explanation isn't that redesigns stopped working — it's that the industry has already fixed most of the glaring usability problems that used to sit in plain sight. A redesign in 2026 is more often refining something that already works than rescuing something broken, which is why a documented case like Baileigh's — a specific, structural failure with a clean before/after test — is a more honest reference point than a generic "redesigns boost conversion 200%" claim with no methodology behind it.

Source	Scope	Evidence Type	Key Finding
Nielsen Norman Group, Baileigh Industrial case study	1 small B2B manufacturer, before/after tree test, n=64 + fresh sample	Documented case study	Findability score rose from 4.0 to 7.4 of 10 (+85%) after a navigation redesign
Nielsen Norman Group, UX Metrics and ROI research	99 case studies across 5 editions since 2006	Aggregate research review	Average KPI improvement from usability redesigns fell from 247% (2006-2008) to 75% (2020)
Baymard Institute, Checkout Usability Research	344 top-grossing US/EU e-commerce sites, 30,000+ elements scored	Benchmark study	Only 2% of sites score "good" on checkout UX; the average site has 32 unimplemented improvements
Clutch, SMB website survey	406 US small business owners, Aug 2025	Self-reported survey	81% have redesigned their website at least once; 90% plan to invest again in the next 12 months

How to Approach Your Own Website Redesign

The pattern across Baileigh's case, NN/g's broader research, and Baymard's benchmarks is that redesigns with measurable results start by diagnosing a specific, testable problem before anyone touches a color palette or a font. Baileigh's team didn't guess that navigation was the issue; they ran a tree test and confirmed it first.

A few things worth taking from that, scaled down to small business size:

• Test before you redesign, not just after. Even an informal version — watching five people try to find something specific on your current site — tells you whether the problem is navigation, trust, speed, or something else, before you spend money fixing the wrong thing.
• Don't confuse a visual refresh with a structural fix. A new color scheme and a new information architecture solve different problems. If people can't find what they need, a prettier version of the same navigation won't fix it.
• Redesign is rarely one-and-done. 81% of small businesses in Clutch's survey have already redesigned once, and 90% expect to invest again within a year — the useful mindset is ongoing refinement, not a single finished project.⁴
• Measure the same thing before and after. Baileigh's team re-ran the identical 8 tasks post-launch. Without that discipline, "it feels better" is the only result you'll have to show for the spend.

Get a Redesign You Can Actually Measure

That last point is where most redesigns quietly fail: nobody defines what "working" means beforehand, and six months later there's no way to know if the redesign helped, hurt, or did nothing. Alphalead builds the creative and the ad and landing-page funnel it feeds into on the same lead-tracking dashboard, so a redesign is measured against your actual cost-per-lead numbers before and after, the same way Baileigh's team measured findability before and after. Most design agencies can't show you that because they never see what happens after the page loads. Talk to our team about scoping a redesign against your real funnel data instead of a portfolio pitch.

Frequently Asked Questions

How do I know if my small business website actually needs a redesign?

Look for a specific, testable symptom rather than a general feeling that the site "looks old" — visitors bouncing from a specific page, people calling to ask questions the site should already answer, or a contact form converting well below what similar businesses report. If you can't name the specific problem, an informal usability test — watching a handful of people try to complete a real task on your current site — is a cheaper first step than a full redesign.

How long does a small business website redesign typically take?

It varies by scope, but the documented example here — a navigation-focused redesign, not a full visual rebuild — went through discovery, a baseline tree test, a card-sort study, implementation, and a follow-up test.¹ A narrow, structural fix is meaningfully faster than a full brand-and-site rebuild, one reason to diagnose the actual problem before scoping the project.

Do website redesigns always improve conversion rates?

No. A redesign that changes visual style without fixing the underlying problem — navigation, trust signals, checkout friction — can leave results flat or hurt them if it disrupts a layout customers had already learned. NN/g's research tracks measured improvements from redesigns that targeted a specific, tested problem, a meaningfully different claim than "redesigns work."² Does a professional logo actually increase sales? covers two real cases, Tropicana and Burberry, where a large brand's identity change didn't play out as a simple win either.

Is a full custom rebuild worth it, or should I redesign on my existing platform?

For most small businesses, no. Clutch's survey found only 12% of small business websites are fully custom-built, and 45% of owners outsource the build within an existing platform like WordPress, Wix, or Shopify.⁴ The measurable gains above come from structural and content decisions, not the underlying platform, so a platform migration is rarely the highest-leverage first move.

Key Takeaways and Next Steps

A useful website redesign starts with a specific, testable problem, not a general sense that the site looks dated. The clearest documented example available, Baileigh Industrial's navigation redesign, shows what that looks like in practice: a measured baseline, a research-backed fix, and a measured result — an 85% improvement in a controlled findability test.¹ The broader research says the same thing at scale: NN/g's numbers show redesign gains shrinking industry-wide as the obvious problems get fixed, and Baymard's benchmarks show even well-resourced e-commerce sites still leave real conversion on the table in flows they assume are finished.^2,3 None of this promises any specific redesign will hit any specific number; it's evidence that the redesigns which measure the problem before and after are the ones worth trusting. This closes out Alphalead's branding and design series — see the full branding guide for how design fits into a broader growth system, how design quality affects ad and landing page conversion rates for the paid-traffic side of this same question, or rebranding a small business without losing existing customers if a name or identity change is part of what you're planning alongside the redesign. If you're deciding whether your site needs a redesign or just a smaller, targeted fix, talk to our team about scoping it against your actual funnel data.

References (4)

1. 1.Nielsen Norman Group — Quantifying UX Improvements: A Case Study (Baileigh Industrial navigation redesign, conducted by Marketade, tree test n=64 baseline + fresh follow-up sample)
2. 2.Nielsen Norman Group — Average UX Improvements Are Shrinking Over Time, 99 case studies across 5 editions of NN/g's UX Metrics and ROI research since 2006
3. 3.Baymard Institute — E-Commerce Checkout Usability Research, 344 top-grossing US/EU sites, 30,000+ checkout elements scored, continuously updated
4. 4.Clutch — No-Code Tools Fuel Website Growth SMB website survey, n=406 US small business owners, Aug 2025