How to read an SEO audit (and decide what to fix first)

You ordered an SEO audit, or a tool produced one for free, and now you are looking at a long report full of warnings, scores and coloured labels. It is easy to feel that everything is broken. Usually it isn't. Most audits mix a few issues that genuinely hold a site back with many small ones that barely matter.

This guide explains how to read a typical SEO audit, what the common sections mean, and how to decide what to fix first. It applies whichever tool or agency produced the report.

First, understand what an audit can and cannot tell you

An SEO audit is a technical and on-page inspection. A crawler visits your pages the way a search engine bot would, reads the HTML and records what it finds: titles, headings, links, status codes, structured data, speed measurements and so on.

That makes an audit good at finding things that are objectively wrong or missing: a page that returns an error, a title that is empty, a sitemap that points to deleted pages. It is much less able to judge whether your content is the best answer for a searcher, or how strong your competitors are. No audit can promise rankings, and you should be wary of any report that does.

Treat the audit as a list of obstacles. Removing obstacles gives your content a fair chance; it does not replace having useful content.

The sections you will usually see

Reports differ in layout, but most cover the same ground.

Crawlability and indexing

This section answers one question: can search engines reach and index the pages you care about? Look for:

• Pages blocked by robots.txt that should be public.
• Pages marked noindex by accident, which is common after a site moves from a staging server.
• Error responses (4xx and 5xx status codes) on pages that should work.
• Redirect chains, where one URL redirects to another that redirects again. Each hop slows crawling and can lose signals.
• Canonical tags that point somewhere unexpected, telling search engines a different page is the "real" version.

Problems here outrank almost everything else. A perfectly written page that is blocked or set to noindex cannot appear in search results at all.

On-page basics

These checks look at each page's core HTML:

• Title tags: missing, duplicated across many pages, or far too long. The title is often what people see as the clickable headline in search results.
• Meta descriptions: missing or duplicated. Search engines don't always use them, but a good one can improve how your result reads.
• Headings: no h1, several competing h1 tags, or headings that don't describe the page.
• Images without alt text, which matters for accessibility and helps search engines understand images.
• Thin pages with very little visible text.

Duplicated titles and descriptions are worth attention when they affect important pages, such as your services or product categories. On a page like a tag archive, they rarely matter.

Links

Audits list broken internal links (links on your site that lead to missing pages) and often broken outbound links too. Broken internal links waste visitors' time and leave crawlers at dead ends. They are usually quick to fix, which makes them good early wins.

Some reports also flag orphan pages, which no other page links to. If a page matters, link to it from somewhere relevant.

Structured data

Structured data (usually JSON-LD) describes your business, products or articles in a machine-readable way. Audits report missing markup and markup with errors. Errors in existing markup are worth fixing, because invalid markup cannot help you. Adding new types is useful but rarely urgent.

Performance and Core Web Vitals

Speed sections often show lab measurements such as Largest Contentful Paint (how quickly the main content appears), Cumulative Layout Shift (how much the page jumps around while loading) and Interaction to Next Paint (how quickly the page responds to taps and clicks). Google uses Core Web Vitals as part of its page experience signals.

Two things help when reading this section:

1. Lab scores vary. Run the same test twice and you may get different numbers. Look for consistent patterns, not one-off results.
2. The suggestions are often more useful than the score. "Serve images in modern formats" or "reduce unused JavaScript" tells you what to do. A score of 62 on its own does not.

Mobile

Most audits check that pages have a viewport tag, readable text and tap targets that aren't crammed together. If your site was built in the last few years with a modern theme, this section is often fine. If it isn't, treat it as important: many of your visitors are likely to be on phones.

How severity labels work, and why you should question them

Tools label issues as errors, warnings and notices, or high, medium and low. These labels are generic. The tool doesn't know which pages earn you money.

A "low" issue on your main service page can matter more than a "high" issue on an old blog post from years ago. So before you act on any label, ask two questions:

1. Which pages are affected? Issues on your home page, service pages, product pages and contact page come first.
2. How many pages are affected? A template problem that touches every page (for example, a missing title on every product) is one fix with a large effect.

A simple order for fixing things

Here is a practical order that works for most small business sites:

1. Anything that stops important pages being indexed. Accidental noindex, robots.txt blocks, wrong canonicals, server errors.
2. Broken internal links and redirect chains on key pages and in your main navigation.
3. Titles and descriptions on your most important pages. Make each one unique and describe what the page offers.
4. Template-level fixes that repair many pages at once, such as heading structure or image alt text in a product template.
5. Performance improvements with clear, specific suggestions, starting with image sizes.
6. Everything else, in batches, when you have time.

Keep a simple list with three columns: the issue, the pages it affects, and whether it is fixed. Many issues will turn out to share one cause.

Common false alarms

Some findings look alarming but are often harmless:

• "Duplicate content" on paginated or filtered pages that already have sensible canonical tags.
• Missing meta descriptions on pages you don't want in search, such as login or cart pages.
• Low word counts on pages that are meant to be short, like a contact page.
• Redirects from http:// to https://, or from www to the bare domain. A single redirect like this is normal and correct; chains of several are the problem.

If you're unsure whether something matters, check whether it affects a page you want people to find through search. If it doesn't, it can usually wait.

Re-check after you fix

After a round of fixes, run the audit again. You're looking for two things: the issues you fixed are gone, and nothing new broke. Changes to themes, plugins and templates often fix one thing and quietly break another.

Search engines also need time to recrawl your pages, so don't expect search results to change the day you publish a fix. For important pages, Google Search Console's URL Inspection tool lets you check how Google sees a page and request indexing after a change.

If your audit says pages have almost no text

One finding deserves a special mention. If an audit reports that pages built with React, Vue or a similar framework have almost no visible text, the crawler may be seeing an empty shell that only fills in after JavaScript runs. We explain why that happens and how to fix it in Why your React site might be invisible to Google.

The short version

Read the indexing section first, then look at which pages each issue affects, not just its colour. Fix whatever blocks important pages, then broken links, then titles and templates. Run the audit again after each round of fixes. A long report doesn't mean a broken site; it means you have a to-do list, and most of it can be done in a sensible order.