Last reviewed: April 24, 2026.
Paying for a redesign before running an audit is one of the most common ways small businesses waste money on their website. The fee quote arrives, the timeline looks exciting, and nobody stops to ask whether the current site is actually broken or just unloved. A structured audit answers that question in a day or two and usually reveals that many of the perceived problems are fixable for less than the cost of a full redesign.
This checklist is built for founders, operators, and marketing leads who want to make an informed redesign decision rather than a reactive one. It covers speed, accessibility, SEO, content, and conversion — all the areas a redesign would touch anyway. The goal is a scored rubric, not a vibe check. When you finish, you should know whether you need a copy refresh, targeted fixes, or a full rebuild.
Quick Summary
- Who this is for: SMB sites, service businesses, local businesses, and lean marketing teams deciding whether to repair or rebuild a website.
- What this audit can tell you: whether the main problems are technical, content-related, trust-related, or conversion-related.
- What this audit cannot tell you: exact redesign ROI, legal compliance in every jurisdiction, or whether a rebrand is strategically necessary.
- What to do if you score high: repair the weak spots before paying for a full redesign.
- What to do if you score low: gather baseline data, identify the failing areas, and use that evidence to scope a rebuild carefully.
Why “Audit First, Redesign Second” Is the Cheaper Path
Most websites that feel outdated have three or four specific problems hidden inside a dozen imagined ones. Founders say “the whole thing needs to be redone” and agencies are happy to scope that. A short audit separates the real problems from the aesthetic discomfort.
The reason to audit first is simple: without a baseline, you cannot tell what the redesign actually improved. Google’s own tools separate lab diagnostics, field data, indexing signals, and page experience signals; treating all of that as one vague “site quality” problem is how budgets get inflated.[1] The audit gives you a before-state, which makes every later redesign conversation more concrete.
2026 Lighthouse and Core Web Vitals Thresholds
Every audit should begin with numeric targets, not opinions. The 2026 practical baseline is still shaped by Google Lighthouse and Core Web Vitals. Miss these and speed, usability, and conversion can all suffer at once.
- Lighthouse Performance ≥ 85
- Lighthouse Accessibility ≥ 90
- Lighthouse Best Practices ≥ 90
- Lighthouse SEO ≥ 90 — useful as a diagnostic score, not a direct ranking factor
- LCP (Largest Contentful Paint) ≤ 2.5s at the 75th percentile
- INP (Interaction to Next Paint) ≤ 200ms — this replaced FID in March 2024[2]
- CLS (Cumulative Layout Shift) ≤ 0.1
Run Lighthouse in an incognito window against your live production URL, not a staging build. Run it three times and take the median score — Lighthouse has real variance between runs. Also check PageSpeed Insights, which can show field data from Chrome users (CrUX) rather than only a one-off lab test.[3]
Search Console: The Free Audit Tool Most Teams Skip
Before any third-party tool, open Google Search Console and check Page indexing, then use URL Inspection on your most important pages. You want four things to be true:
- Zero unresolved errors affecting important URLs in Page indexing.
- All primary pages show an indexed status when inspected individually.
- Every canonical URL resolves to a 200 response.
- Structured data validation is clean — no invalid states on service and product schemas, and warnings reviewed in context.
If Search Console shows dozens of “Crawled — currently not indexed” pages, you may have a thin-content, duplicate-content, or crawl-priority problem that no visual redesign will fix. That is a content-strategy problem. Avoid using a Google site: search as your indexing test; it can be useful for rough discovery, but Search Console is the better source of truth for your own property.[4]
The 27-Point Checklist
Score each item Green, Amber, or Red. Green means the item meets the current target or is clearly healthy for your site type. Amber means it is close but needs review. Red means it is failing and visibly hurting visitors, rankings, measurement, or sales conversations.
| # | Area | Check |
|---|---|---|
| 1 | Speed | Lighthouse Performance ≥ 85 on mobile |
| 2 | Speed | LCP ≤ 2.5s on top 5 pages |
| 3 | Speed | INP ≤ 200ms on primary CTAs |
| 4 | Speed | CLS ≤ 0.1 during page load |
| 5 | Speed | Images served in modern formats where appropriate (WebP/AVIF) |
| 6 | Speed | Render-blocking JS deferred, reduced, or justified |
| 7 | Accessibility | Lighthouse Accessibility ≥ 90 |
| 8 | Accessibility | Informative images have descriptive alt text; decorative images use empty alt |
| 9 | Accessibility | Color contrast meets WCAG AA (4.5:1 for normal body text) |
| 10 | Accessibility | Forms have visible labels and usable error states |
| 11 | SEO | Lighthouse SEO ≥ 90 as a diagnostic check |
| 12 | SEO | Search Console Page indexing has no unresolved errors on priority URLs |
| 13 | SEO | Every indexable page has a unique title and meta description |
| 14 | SEO | Primary pages pass URL Inspection for indexing and canonical status |
| 15 | SEO | Structured data valid for relevant services, products, reviews, or local business details |
| 16 | SEO | XML sitemap submitted and up to date |
| 17 | Content | Homepage hero answers “what, for whom, what next” in 5 seconds |
| 18 | Content | Service pages are substantial enough to explain offer, proof, process, and fit |
| 19 | Content | About page includes credibility signals such as credentials, experience, team, or founder detail |
| 20 | Content | Contact page shows the contact methods customers actually need; address and hours included when relevant |
| 21 | Trust | Testimonials include names, companies, roles, or other credible attribution where permitted |
| 22 | Trust | Case studies show outcomes, not just screenshots |
| 23 | Trust | Privacy policy, terms, and cookie notice match the site’s jurisdiction and data collection |
| 24 | Conversion | Primary CTA visible or quickly reachable in every key viewport |
| 25 | Conversion | Forms ask only for fields needed at that stage of the buyer journey |
| 26 | Conversion | GA4 conversion events firing for form, call, email, and other primary actions |
| 27 | Conversion | Thank-you page or confirmation flow in place |
How to Check Each Section
Speed: Run Lighthouse and PageSpeed Insights on the homepage, the highest-traffic service page, and the main conversion page. Speed matters because slow pages create friction before copy or design has a chance to work.
Accessibility: Use Lighthouse, keyboard navigation, form testing, and a contrast checker. Accessibility is not just a compliance exercise; it catches basic usability problems that affect every visitor.
SEO: Use Search Console Page indexing, URL Inspection, your sitemap, and structured data testing. SEO checks matter because a redesign that damages indexing or canonicals can look beautiful while quietly losing traffic.
Content: Read the homepage, service pages, about page, and contact page as if you are a first-time buyer. Content matters because unclear positioning is often misdiagnosed as a visual-design problem.
Trust: Look for proof that a real business has delivered real outcomes. Trust matters because visitors need evidence before they commit time, money, or personal information.
Conversion: Test the CTA path on mobile and desktop, then verify the events in GA4. Conversion matters because a redesign without working measurement is just a new set of guesses.
The Decision Rubric: What Your Score Actually Means
Once you have scored each of the 27 items, count Greens. The bands below are internal SMB heuristics, not universal rules. They are most useful for service businesses, local companies, consultants, and lean B2B sites where the website has a small number of primary conversion paths.
- 22+ Greens → Repair path first. You likely need copy cleanup and a few technical fixes. Scope small repairs before considering a redesign.
- 15–21 Greens → Targeted fixes plus messaging rebuild. You likely need a partial rebuild focused on the failing areas plus a rewrite of hero, services, and CTAs.
- 14 or fewer Greens → A full redesign may be justified. Before scoping, verify this with a second opinion and baseline data — vendors have a financial incentive to call broad problems a redesign.
Any item you marked “unclear” during the audit should trigger diagnostic research, not redesign spend. The mistake is treating uncertainty as a reason to rebuild.
The Traps That Inflate Redesign Budgets
Audits keep budgets honest, but the checklist can still get hijacked if you do not watch for these patterns.
- “It feels dated.” That is aesthetic, not operational. Dated visuals rarely score Red on the 27-point list. Check the numbers before agreeing a redesign is needed.
- “Our competitor has a flashier site.” Competitor envy is one of the most expensive briefs in this industry. If your score is high, a flashier competitor may not be why you are losing deals.
- “The CMS is outdated.” Sometimes true — but often a migration, not a redesign, is the real need. Ask the specific question: does the platform actually block new features?
- “The analytics are broken.” Fix the analytics first. You cannot justify a redesign with missing baseline data.
What to Do With the Audit Result
Once you have the score and decision path, the next step is writing a one-page brief that states the current score, the target score, the three highest-impact fixes, and the budget ceiling. That brief is how you prevent scope creep when agencies respond.
For teams running on WebsiteBuilder, many high-scoring sites can be improved in-platform without an external redesign engagement — copy rewrites, CTA moves, form-field reductions, and image-format upgrades are all practical repair work. That is usually the cheapest redesign path to test before paying for a broader rebuild.
FAQ
How long does the 27-point audit take?
With Lighthouse, Search Console, GA4, and a live browser tab open, most founders finish in two to three hours. Budget a half day if you are doing it carefully and documenting the results.
Do I need a tool subscription to run this audit?
No. Lighthouse, PageSpeed Insights, Search Console, and GA4 are all free. Paid tools like Ahrefs or Screaming Frog add depth but are not required for a 27-point check.
What if my score is borderline, like 21 Greens?
Run the audit on your top three pages individually — homepage, primary service page, and contact page. If any of those score below 18, treat it as the 15–21 band. The top three pages usually do most of the conversion work.
Should I retake the audit after redesign or fix work?
Yes. Re-audit 30 days after launch to confirm the fixes stuck. Also re-audit quarterly as Google adjusts reporting, metrics, and documentation — INP replaced FID in 2024 and future measurement priorities may shift again.
Is a redesign ever justified without an audit?
Rarely. The clearest case is a rebrand that changes the business name, positioning, or target customer so materially that no part of the current site still applies. Even then, run the audit first to preserve the technical decisions that are working.
Sources
- Google Search Central, Page experience and Core Web Vitals documentation: https://developers.google.com/search/docs/appearance/page-experience
- Google Search Central, INP replacing FID as a Core Web Vital in March 2024: https://developers.google.com/search/blog/2023/05/introducing-inp
- Google PageSpeed Insights and Lighthouse documentation: https://pagespeed.web.dev/ and https://developer.chrome.com/docs/lighthouse/performance/performance-scoring
- Google Search Console, Page indexing and URL Inspection documentation: https://support.google.com/webmasters/answer/7440203 and https://support.google.com/webmasters/answer/9012289
- W3C Web Content Accessibility Guidelines contrast and non-text content guidance: https://www.w3.org/WAI/WCAG22/Understanding/contrast-minimum.html and https://www.w3.org/WAI/WCAG22/Understanding/non-text-content.html