How to Do an SEO Audit: 50 Checkpoints That Cost You Rankings

Short answer
How to do an SEO audit, and in what order should you work?
An SEO audit is a structured check of the problems that stop search engines from finding, understanding, and ranking your pages, and that stop visitors from having a good experience. You work in a fixed order: crawling and indexing first, then duplicate URLs, then content, links, and page experience. Also, this guide turns that order into 50 checkpoints.
The order matters. In practice, if a page never reaches the index, debating its content quality gets you nowhere. So walk the list from top to bottom, and write yes, no, or unknown next to every item.
Also, the facts here rest on Google Search Central and web.dev documents. Google's own SEO starter guide forms the backbone of this audit. Details can change, so confirm critical decisions in the source document.
How to do an SEO audit: what should you prepare first?
Before you start, settle three things: which pages matter, which searches you want to appear for, and which tools you have. In practice, without those, the checklist turns into a random dump of errors.
First, list your important pages. For example, the home page, your service or product pages, and your top traffic articles are enough. Then check your Google Search Console access. Also, without it, half of the audit rests on guesses.
- Important page list: a few dozen URLs at most.
- Access: Search Console, analytics, and the site admin panel.
- Reference: earlier reports and recent site changes.
For a quick first look, run your home page through the free seo checker. So the findings show which block deserves priority. Verifying your measurement at this stage also pays off, because you need trustworthy data to read any traffic drop.
If you only want to review one page fast, our guide to a 10 minute SEO check of a page is enough to start. Also, do not try to finish the audit in one sitting. Even a small site needs separate sessions for crawling, content, and performance.
If the site is too large to handle alone, an experienced seo consulting team speeds up the preparation and the prioritization. For small sites, this guide may be enough by itself.
Which 10 crawling and indexing checkpoints should you check?
The first block asks whether Google can find your pages and whether it indexes them. In practice, if one of these items breaks, every other effort goes to waste. Google's robots.txt guide and its page on blocking indexing form the base of this section.
| Checkpoint | How to check | Why it matters |
|---|---|---|
| 1. Key pages appear in the index | Use a site: search and Search Console | No index entry means no ranking |
| 2. A sitemap exists | Open the file and check Search Console | Speeds up discovery |
| 3. Sitemap lists only preferred HTTPS URLs | Sample the listed addresses | Avoids conflicting signals |
| 4. robots.txt does not block key pages | Read the file and its rules | Blocked pages never get crawled |
| 5. robots.txt does not act as a hiding tool | Look for noindex on pages you want hidden | Blocked URLs can still appear in results |
| 6. noindex sits only on intended pages | Check the meta tag and the response header | A wrong noindex removes the page |
| 7. noindex pages stay open to crawling | Check for a robots.txt clash | A blocked page hides the rule |
| 8. CSS and JavaScript stay open | Read the URL Inspection output | The page may look different to Google |
| 9. Google's view matches the user's view | Compare in URL Inspection | Reveals hidden content risk |
| 10. Key pages get internal links | List orphan pages | Google finds pages through links |
In short, the result of this block is the ground floor of everything else. Rescue pages that cannot be crawled or indexed first, then optimize them.
Also add proof to each row: a screenshot, an address, or a date. For example, a yes without proof is a guess you must verify again later. Keep a separate list of pages that carry noindex on purpose.
Which 8 duplicate URL and canonical checkpoints should you review?
When the same content opens at several addresses, signals split. Google's page on consolidating duplicate URLs ranks the methods from strong to weak: redirects, the rel canonical annotation, and the sitemap. So the table turns the principles of that page into audit items.
| Checkpoint | How to check | Why it matters |
|---|---|---|
| 11. Each piece of content has one preferred address | Try www, HTTP, and slash variants | Signals gather on one URL |
| 12. Duplicates merge by redirect or canonical | Sample parameter and repeat URLs | Cuts duplicate waste |
| 13. Canonical page points to itself | Read the source code | States a clear preference |
| 14. Canonical uses an absolute address | Look for relative paths | Prevents misreading |
| 15. Signals stay consistent | Compare sitemap, tag, and internal links | Conflicts confuse Google |
| 16. No wrong tool stands in for canonical | Check robots.txt, noindex, and removal tool use | The page may drop out entirely |
| 17. Permanent redirects work correctly | Test the redirect type | A permanent redirect sends a strong signal |
| 18. URLs use readable words | Find random ID addresses | Users and Google understand them |
For the details, read our guide to the canonical tag. To debug problems, see how to detect canonical issues in Search Console.
In short, the key principle of this block is consistency. If you prefer one address, use that same address in internal links, in the sitemap, and in the canonical tag. Even a small www difference can send Google mixed signals. So put the three side by side on a few random URLs, and most conflicts show up.
Which 6 title and snippet checkpoints should you audit?
Titles and descriptions give the first impression in results. According to the starter guide, every page needs a unique, clear, and concise title element. Also, a description summarizes the most relevant points of the page. However, Google may sometimes choose its own snippet.
| Checkpoint | How to check | Why it matters |
|---|---|---|
| 19. Every page has a title element | List gaps with a crawl tool | The title is a core signal |
| 20. Titles stay unique | Group repeated titles | Pages stay distinct |
| 21. Titles run short, clear, and accurate | Compare title with content | Meets the click expectation |
| 22. A description exists and fits the page | Find missing and repeated descriptions | Improves snippet quality |
| 23. No meta keywords and no keyword stuffing | Clean up stuffed fields | Ranking value stays low |
| 24. Headings describe the sections | Read the heading outline | Helps readers and screen readers |
While you write, lean on our guide to writing a meta title and description. To preview the result, you can try the free google serp preview tool.
Which 6 content quality checkpoints should you check?
However, a content audit does not mean counting words. The starter guide says word count targets have no ranking value by themselves. The real standard is original, well organized, and current content that you write for people first.
| Checkpoint | How to check | Why it matters |
|---|---|---|
| 25. Content is original | Scan for overlap with other pages | Valuable content stands out |
| 26. Text reads well and has no errors | Review spelling and paragraph flow | Raises user trust |
| 27. Content stays current | List outdated pages | Old facts hurt trust |
| 28. Pages answer the search intent | Compare query with page promise | Intent match keeps clicks |
| 29. Ads and interstitials do not intrude | Open the page on a phone | Content stays reachable |
| 30. No pointless repetition | Merge similar pages | Duplicates are no penalty but waste effort |
For duplicate and thin content, see our post on duplicate content. If you plan a new article, review your existing pages first, because strengthening an old page often beats writing a new one.
In a content audit, give every page one decision: keep, update, merge, or remove. In practice, pages that pile up without a decision tire both users and crawlers later. A small table keeps those decisions simple.
Which 6 link and site structure checkpoints should you review?
Google mostly discovers pages by following links. So internal linking matters for both discovery and meaning. Anchor text should tell readers what the target page is about.
| Checkpoint | How to check | Why it matters |
|---|---|---|
| 31. Anchor text describes the target | Search for generic phrases like click here | Explains the destination |
| 32. Untrusted links carry nofollow | Check user generated areas | Protects trust signals |
| 33. Key links stay crawlable | Confirm they work as real links | Enables discovery |
| 34. Similar content groups in folders | Review the URL structure | Helps on large sites |
| 35. Breadcrumb markup works | Validate the structured data | Can show the path in results |
| 36. Multilingual sites use reciprocal hreflang | Check the language pairs | The right language version surfaces |
For linking strategy, our guide to internal linking for SEO gives practical examples. Finally, make sure every important page gets at least one meaningful link.
Which 6 image, video, and structured data checkpoints should you audit?
Images and structured data also shape how well Google understands a page and whether it qualifies for rich results. The starter guide recommends sharp images near related text and descriptive alt text. Valid structured data makes a page eligible for features such as review stars or carousels.
| Checkpoint | How to check | Why it matters |
|---|---|---|
| 37. Images look sharp and sit near text | Browse key pages by eye | Strengthens context |
| 38. Alt text describes the image | Find empty and generic alt text | Helps accessibility and meaning |
| 39. Video pages stand alone | Check that each video has its own page | Makes video discovery easier |
| 40. Structured data validates | Test with a validation tool | Broken markup loses value |
| 41. Data type fits the page | Check the target feature in the docs | Decides eligibility |
| 42. Image and video sitemap entries exist | Add them to the sitemap if needed | Supports media discovery |
For details, see our guides to alt text and schema markup. Before you add structured data, make sure the page content really reflects the markup.
Which 8 page experience and security checkpoints should you check?
This block measures speed and trust. web.dev recommends three Core Web Vitals metrics: LCP for loading, INP for interactivity, and CLS for visual stability. Confirm the current thresholds in the web.dev document, because thresholds can change.
| Checkpoint | How to check | Why it matters |
|---|---|---|
| 43. LCP sits in the good range | Look at field data | Measures perceived loading |
| 44. INP sits in the good range | Look at field data | Measures interaction delay |
| 45. CLS sits in the good range | Find shifting elements | Measures visual stability |
| 46. Measurement uses field data, split by mobile and desktop | Judge at the 75th percentile | That reflects real user experience |
| 47. Lab tests run during development | Run Lighthouse | Gives fast feedback |
| 48. Pages load over HTTPS with a valid certificate | Watch for browser warnings | A certificate issue pushes toward HTTP |
| 49. No HTTPS to HTTP redirects | Check the sitemap and hreflang addresses | Prevents the wrong version winning |
| 50. Index status gets regular monitoring | Check the Search Console index report | Catches problems early |
To understand the metrics, read the section on field data below. If you find unindexed pages, follow the steps in how to find unindexed pages in Search Console.
How to do an SEO audit with an automatic first scan?
Before you audit by hand, an automatic first scan gives direction. Open the free analysis page, enter your main address, and match the findings to the items in this guide. That way you see quickly which block needs priority.
For example, scan your home page and your most important service page separately. Then compare the title, description, and structure findings with items 19 to 24. Also, if you find missing or repeated titles, add them to your list.
In practice, tools are guides, not judges. You must decide for yourself whether a warning really harms your site. Also save the free seo analysis page so you can repeat the same scan after each change.
If you lack a sitemap or a robots.txt file, you can generate them with the xml sitemap generator and the robots.txt generator. So read every rule once before you publish the generated file.
What do you check in Search Console during an audit?
Search Console is the most reliable data source for an audit, because it shows how Google sees your site. For example, the starter guide recommends checking whether your site appears in the index and whether technical barriers exist. The URL Inspection tool stands out for this job.
With URL Inspection you compare how Google sees an address with how a user sees it. That catches blocked resources and hidden content. So this comparison is the proof for items 8 and 9 on the list.
Also read the index report for the reasons pages stay out. However, each status means something different. Some point to a deliberate noindex decision, and others point to a real problem. Also, read each status on its own to tell them apart.
- Index report: excluded pages and their reasons.
- URL Inspection: the HTML Google saw and any blockers.
- Sitemap report: submitted addresses and readability.
Screen names change over time. So instead of a fixed menu path, look for the matching section in the panel.
How do you prioritize the findings of an SEO audit?
In practice, an audit can produce hundreds of findings. You cannot fix them all at once, and you do not need to. Prioritize on an impact and effort axis. First come the problems that block indexing, then content and experience.
| Priority | Finding type | Example | When |
|---|---|---|---|
| 1. Urgent | Blocks indexing or crawling | Wrong noindex, blocking robots.txt | Right away |
| 2. High | Signal conflict | Canonical and sitemap mismatch | Within days |
| 3. Medium | Snippet and content | Repeated title, missing description | Within weeks |
| 4. Medium | Page experience | Weak LCP or INP | Planned sprint |
| 5. Low | Fine tuning | Alt text polish | Routine upkeep |
So give each finding an owner, a date, and an expected impact. That turns the audit from a list into a work plan. For an item with unclear impact, a small test first makes sense.
Why do robots.txt and noindex clash?
In practice, this is the most common clash. According to Google's documentation, robots.txt is not a way to keep a page out of Google. To hide a page, you use noindex or password protection. Also, a blocked page can still show in results without a description when other sites link to it.
Here is the subtle part. For noindex to work, Googlebot must be able to reach the page. However, if robots.txt blocks the page, Googlebot never sees the noindex rule. So using both measures together cancels each one out.
- First, confirm that robots.txt does not block the page.
- Then check that the noindex tag or header exists, using URL Inspection.
- Finally, wait for Google to recrawl the page.
Also, if you need fast removal, read Google's removal documentation. To see the difference in depth, read our comparison of noindex, nofollow, and robots.txt. For the rules themselves, our post on common robots.txt mistakes helps too.
How do you choose between a canonical tag and a redirect?
Both merge duplicate addresses, but they fit different cases, so choose by situation. A permanent redirect suits the moment you retire a page and also want to send the user to the new one. A canonical tag suits the case where both pages stay reachable and you only want to name the preferred address.
Google's document ranks redirects as a strong signal, the canonical annotation as a medium one, and the sitemap as a weaker one. Combining methods strengthens the signal. However, conflicting signals break that effect.
- If a page moved for good, use a permanent redirect.
- If you need both versions, such as a filtered list page, use a canonical tag.
- In both cases, point internal links to the preferred address.
So do not use noindex for canonicalization. The document says it removes the page from search entirely. Also set redirects from HTTP to HTTPS, never the other way round.
How do you read Core Web Vitals results?
According to web.dev, all three metrics are primarily field metrics. That means they capture real user experience and show up in tools that draw on the Chrome User Experience Report. In practice, a page passes when it meets the recommended targets at the 75th percentile for all three metrics. You judge mobile and desktop separately.
Lab tests help during development. However, Lighthouse cannot measure INP directly, because INP needs real user input. So total blocking time serves as a lab proxy for INP.
In practice, look at field data first. However, if there is not enough data, a page may show no result. Read that case with the method we describe in our guide to Core Web Vitals.
When you find a slow page, look for the pattern, not one cause. Large images, late scripts, and ad slots without reserved space are common shared causes.
Which mistakes do people make when they audit SEO?
However, the audit itself can create errors. The most frequent ones are filling the list mechanically and accepting tool output as it stands. So the list below sums up these traps.
- Calling a tool warning critical without checking your context.
- Jumping to content work before you solve indexing problems.
- Giving every page the same weight.
- Reporting findings with no owner and no date.
- Forgetting to measure again after a change.
- Spending time on low value fields such as meta keywords.
Another mistake is treating a fix as a guarantee. No audit promises a ranking gain. In practice, an audit removes obstacles, while content, competition, and time decide the ranking.
Finally, do not treat the audit as a one-off job. Problems return as the site changes, so regular repetition is a must. Plugin updates, template changes, and new content types can quietly break old fixes. Rerun the related items after every big change.
What should the report of an SEO audit look like?
So a good audit report is short, ordered, and ready for action. The reader should see what to do on the first page. In practice, a long error dump delays decisions. So present findings with priority, owner, and date.
- Summary: the top three findings and their expected impact.
- Finding list: checkpoint number, status, and evidence.
- Priority table: urgent, high, medium, low.
- Action plan: who does what, by when, and how you will measure it.
To read such a report, see our post on how to read an SEO report. For another 50 point audit, our CRO audit guide uses a similar structure.
Always put a verification date on the report. When you run the same list three months later, you see progress in concrete terms.
What should you check after a migration or redesign?
A redesign is the moment when an audit pays off most. Addresses change, templates refresh, and old signals can vanish. In the first week after a move, run the indexing, redirect, and canonical items again. For a detailed list, see our website migration SEO checklist.
First confirm that old addresses redirect to the new ones permanently. Google's document describes permanent redirects as a strong signal. Then check that the sitemap lists the new addresses.
Third, make sure titles, descriptions, and structured data survive in the new template. A template change can erase these fields without any warning. Last, compare the Core Web Vitals results, because a new design also changes speed.
- Test redirects from old addresses.
- Recheck the sitemap and robots.txt file.
- Compare titles, descriptions, and markup.
- Watch speed and the index report for a few weeks.
What routine should follow the audit?
The value of an audit shows in the routine that follows. A monthly index and error check plus a full list run every three months suits most sites. Add extra checks before and after any major site change.
The index report in Search Console helps you watch pages where Google found a noindex rule. Google's sitemap document reminds you that a sitemap does not guarantee that every item gets crawled or indexed. So a sitemap alone is not enough, and you need to monitor it.
For a small site, roughly 500 pages or fewer, a sitemap may not be required. Still, most sites benefit from one. Confirm this in the sitemap documentation and check whether your CMS already makes a sitemap. For more detail, see our guide to the XML sitemap.
- Weekly: skim Search Console alerts.
- Monthly: check index, speed, and title changes.
- Quarterly: run the 50 point list from the top.
Should you audit SEO yourself, and when should you get help?
On a small site, you can do most of the audit yourself with this list. The hard part is usually prioritization and technical reading. Deciding whether a finding is a real problem takes experience.
At that point, professional seo consulting services give you a pair of eyes that turns findings into a work plan. As Talha Aslan and our team, we handle the audit, the fixes, and the measurement afterwards together. We do not promise rankings, because such a promise would not be honest.
Whether you work alone or with help, keep the checklist in writing and date every finding. Regular audits catch small problems before they grow. At the end of each audit, measure again and carry the result into the next round.
Core guide: What Is Digital Marketing? Its types, examples and how it is done, on one page with an interactive film that travels through the brain.



