TL;DR: Good SEO audit services don’t hand you a tool export. They tell you which problems affect the pages that matter, what to fix first, who owns each fix and how to check it worked. A one-off SaaS technical audit typically costs $2,500–$10,000, depending on hosts, templates and depth. The biggest red flags are a score with no diagnosis, every issue marked high priority, and no plan to verify fixes. If nobody is available to implement the findings, don’t buy an audit yet.
Modern crawlers can find thousands of issues on almost any website in an afternoon. That is the easy part, and it is why so many audits read like a printout.
The hard part is deciding which of those issues matter. An audit earns its fee when it turns a pile of warnings into a short, ordered list that a team can ship.
This guide covers what a real audit includes, what it costs, which type you need, and how to tell a useful one from an expensive PDF. It is written for founders and marketing leaders at B2B SaaS and AI companies, though most of it applies to any site.
SEO Audit Template
54 checks for Google and AI search that score your site and rank your fixes by impact and effort. Includes an AI prompt tracker and a worked example.
Download the free templateWhat is an SEO audit?
An SEO audit is a diagnosis of what is stopping a website from being crawled, understood, ranked, cited and converted, with the problems placed in the order they should be fixed.
The second half of that sentence is the part that gets skipped. A list of issues is data. A diagnosis explains which issues are real, which share a root cause, and which ones affect pages that bring in revenue.
An audit is also a one-time diagnostic product. It is different from ongoing SEO, where the same team implements, publishes and measures every month.
What does a real SEO audit include?
A real audit starts with the business, then works through access, structure, content, authority and AI visibility. This is the sequence we follow, with depth adjusted to the site.
- Baseline. We document what you sell, to whom, in which markets, and which pages generate leads. We record current traffic, rankings and conversions where data exists.
- Technical access and indexation. Robots.txt on every host, sitemaps, noindex directives, canonicals, redirects, status codes, JavaScript rendering, and raw HTML against rendered HTML.
- Architecture and internal links. Orphaned pages, weak hubs, blog posts that never link to commercial pages, and links pointing at redirects.
- Content and on-page targeting. Search intent, cannibalisation, thin templates, and missing comparison, alternative and use-case pages.
- Authority and competitor gaps. Referring domains, authority reaching commercial pages, and searches where competitors appear and you don’t.
- AI-search baseline. Whether assistants mention, cite and describe your brand correctly for a fixed set of buyer prompts.
For SaaS companies, step two usually takes the longest. The marketing site, app, docs, blog and help centre often sit on different platforms, and each host can have its own robots, sitemap and rendering problems. Our guide to technical SEO services for SaaS covers that layer in detail.
What should you receive at the end?
A useful audit gives you:
- A prioritised findings document with evidence
- A URL-level or template-level action list
- An owner, an effort rating and an impact rating for each action
- A 90-day roadmap
- A measurement plan
- A walkthrough, so your team understands what to do next
What does the auditor need from you?
At minimum: the site and its subdomains, target markets, priority products, main competitors, and access to Search Console and analytics if you have them.
Missing access should reduce the certainty of the diagnosis, not stop the audit. Without Search Console we can still crawl the public site, review rankings and baseline AI visibility. We can’t confidently connect findings to historical clicks or pipeline.
How much does an SEO audit cost?
A one-off SaaS technical SEO audit typically costs $2,500–$10,000. That is the range we publish, and we quote each audit on complexity, not on a fixed menu.
Our monthly search-visibility programmes, from $2,500, $5,000 and $8,000 a month, are a separate thing. They cover ongoing content, technical work, authority and reporting, and should not be read as audit prices.
What moves the price?
URL count is only one factor. These matter more:
- Hosts and subdomains.
www,app,docs,helpandblogare five sites, not one. - Templates. Fifty templates take longer than 5,000 pages on one template.
- JavaScript rendering. Raw and rendered HTML have to be compared.
- Markets and languages.
- Depth. Whether content, authority and AI-search baselining are in scope.
- Follow-through. Whether you want developer tickets, implementation support and re-checks.
What does the market charge?
Survey data gives a reference point. In Ahrefs’ poll of 439 SEO providers, the most common per-project fee was $2,501–$5,000. For consultants it was $5,001–$10,000, and for freelancers $101–$250. The most common hourly rate was $100–$150.
Those figures describe SEO projects in general, not audits alone, so treat them as context.
Is a free audit worth anything?
Yes, as a screening tool. I’ve sent many free reports myself. A free report can show that a problem exists: a slow page, a broken link, a missing title.
It rarely includes Search Console analysis, business context, manual validation, URL-level recommendations or any verification. A free audit can tell you where to look. A paid audit should tell you what matters, what to do first and how to confirm it worked.
When is an audit a waste of money?
- The site has 10 to 15 pages and little search demand
- The site is about to be replaced, and the audit won’t inform the migration
- Nobody is available to implement the findings
- You only want a score
The hidden cost of a cheap audit is developer time. Eighty vague tickets cost more to work through than twelve clear ones.
Want to run a first pass yourself before you hire anyone? Our free template covers the same ground as a professional audit, including AI search, and ranks the fixes for you.
Get the free SEO Audit Template
- 54 checks across technical, on-page, content, authority, AI search and measurement
- A scored summary and a 25-item action plan that ranks itself
- An AI prompt tracker for ChatGPT, Perplexity, Gemini, AI Overviews, Claude and Copilot
Free. Excel + Google Sheets. No spam, unsubscribe any time. See the full template page.
Your template is ready
Download it below. We've also sent the details to our team, so reply to any email from us if you'd like a hand filling it in.
If you are budgeting for ongoing work after the audit, our guide to how much SEO costs explains retainers, project fees and what each budget level buys.
Which type of SEO audit do you need?
Match the audit to the problem. A full audit is not automatically the best one.
| Your situation | Start with |
|---|---|
| Traffic dropped suddenly | A technical and search-performance diagnostic |
| Publishing a lot, nothing happens | A content, intent and architecture audit |
| About to redesign or migrate | A migration audit, before development finishes |
| Ranking, but no leads | A search-intent and conversion-path review |
| Invisible in ChatGPT or Perplexity | An AI-search audit with a crawler-access review |
| New site, no traffic | A foundational technical and strategy review |
Two of these deserve a warning.
A traffic drop is not always an algorithm update. We have seen public pages lose visibility because of the site’s own configuration. Check recent releases, canonicals, robots rules and rendering before blaming Google. Our guide to running a traffic loss audit sets out the order we check things in.
A migration audit after launch is too late. The valuable work is the URL map, redirect rules and template requirements agreed before the new site goes live.
Full audits are periodic. Monitoring is continuous: indexation, redirects, rankings, organic conversions and release-related regressions should be watched between audits.
What are the biggest SEO audit red flags?
The worst audits are rarely wrong. They just fail to say what matters, what happens first and who acts.
Red flags before you buy
- A fixed fee quoted without seeing the site
- Promised rankings or a guaranteed traffic increase
- A “complete audit” of a complex site in 24 hours
- No questions about Search Console, analytics or which pages make money
- No sample finding to show you
- An audit that is clearly a pitch for a retainer
A fixed price or a fast turnaround is fine for a small, simple site. The warning sign is either one offered without understanding the work.
Red flags in the report
- The tool export. Ninety pages of warnings with no decision framework.
- A score instead of a diagnosis. “Your site scored 62 out of 100” describes a tool’s weighting, not your site.
- Everything is high priority. A blocked pricing page and a missing image dimension get the same label.
- No affected URLs. “Improve internal linking” names no pages.
- No evidence. Nothing shows how a finding was verified.
- Generic priorities. Crawl-budget work on a 300-page site, or a perfect Lighthouse score while key pages aren’t indexed.
- No owner and no verification plan.
Things that are not red flags
- A short report. Twenty pages of evidence beat 150 pages of export.
- “This doesn’t need fixing.” That is a good sign. You don’t need every warning turned into paid work.
- Sampling. Large sites should be reviewed by template and segment, not URL by URL.
If an audit produces 100 equally urgent actions, it has avoided the hardest part of the work: deciding what matters.
What does a good audit finding look like?
A good finding answers eight questions: what is wrong, why it matters, which URLs or templates, how it was verified, what should change, who owns it, how hard it is, and how you’ll know it’s fixed.
Here is an illustrative finding, based on an anonymised pattern we have seen. It is not a verbatim client ticket.
Finding: public pages canonicalise to a logged-in dashboard URL
Priority: critical. Owner: engineering, with SEO verification. Effort: medium.
- What is wrong. A group of public tool pages carries a canonical tag pointing at a logged-in dashboard URL instead of the page’s own address.
- Why it matters. Search engines are told the preferred version is a page they can’t access. Signals consolidate to the wrong URL, and the public pages can drop out of the index. The result can look like an algorithm hit.
- How it was verified. Crawl of the URL group, canonical checked in source HTML, target tested without authentication, representative URLs inspected in Search Console, sitemap and internal links reviewed.
- What should change. Self-referencing canonicals on public pages, dashboard URLs removed from the canonical logic and the sitemap, and the template fixed so new pages don’t inherit the fault.
- How to confirm it. Re-crawl the group, confirm each canonical resolves to a public 200 page, re-inspect sample URLs, then monitor indexation and impressions.
Fixing it removes a constraint. It is not a promise that all lost traffic returns.
Weak wording versus actionable wording
| Weak | Actionable |
|---|---|
| Improve internal linking. Priority: high. | The pricing page gets no contextual links from the five strongest product and comparison pages. Add them with descriptive anchors. Owner: content. Effort: low. Re-crawl to confirm. |
| Improve crawlability. | 134 of 155 sitemap URLs are leftover theme-demo pages. Remove them, confirm the rest are canonical and return 200, resubmit. Owner: engineering. Effort: low. |
| Fix missing alt tags. | About 400 images lack alt text. Update the meaningful ones for accessibility, and schedule this after indexation and rendering blockers. Owner: content. Effort: medium. |
The sitemap example is real, anonymised. So is another site we reviewed with an empty sitemap, a broken canonical and 27 orphaned articles. No health score would have explained either.
How findings should be ordered
We rank by four questions. Can the page be found? Can it be rendered and understood? Is it commercially valuable? Can the team realistically ship the fix?
| Priority | Illustrative finding | Why it sits here |
|---|---|---|
| 1 | Pricing page carries a noindex | A revenue page can’t be indexed |
| 2 | Public pages canonicalise to a dashboard | Signals go to the wrong URL |
| 3 | Key copy and links appear only after JavaScript runs | Some crawlers get incomplete content |
| 4 | Sitemap lists demo and redirected URLs | Poor signals about which pages matter |
| 5 | 27 orphaned articles | Useful content is hard to discover |
| 6 | 400 images lack alt text | Worth doing, lower urgency |
What should you ask before hiring an auditor?
Send these ten questions to any provider. The answers will tell you most of what you need.
- What is included in the scope, and what is excluded?
- How many URLs, templates, hosts and subdomains will you review?
- What access do you need, and what can you do without it?
- Can I see one complete sample finding?
- How do you separate critical blockers from tool warnings?
- Will recommendations be mapped to URLs or templates?
- Who implements the fixes, and who verifies them?
- How will success be measured?
- What is the timeline, and what does it depend on?
- Who does the analysis: a senior SEO, or software and a junior?
Question four matters most. One finding, shown in full, reveals more than any sales deck.
If you’d rather build this capability in-house, a small stable site with one host can often be audited with a crawler, Search Console and time. An outside auditor earns the fee when the site is complex, the cause is unclear, or a migration is at stake. Our guide to SEO consulting services covers when that trade-off tips.
What does a 2026 audit add for AI search?
A current audit measures whether AI assistants mention, cite and correctly describe your brand, alongside rankings and indexation. For a B2B SaaS company, that belongs inside the audit, not in a separate product.
How to baseline AI visibility
One answer from one chatbot proves nothing, because answers vary by location, wording, date and model. A workable baseline is 25 to 30 buyer prompts, across the assistants your buyers use, run three times each, and repeated later with the same prompts.
Track five numbers: mention rate, citation rate, share of recommendations, where the citations come from, and how often the description of your company is accurate. Our free AI Search Strategy Checklist includes a prompt worksheet for this.
Should you allow AI crawlers?
Separate search crawlers from training crawlers. OpenAI documents OAI-SearchBot for ChatGPT search and GPTBot for training, and says each can be controlled independently. Anthropic lists three bots, and Perplexity states PerplexityBot is for search results, not model training.
My default for a public commercial site: allow the search crawlers where you want visibility, and decide separately about training crawlers. Blocking is reasonable for confidential, paywalled or licensed content. Check the CDN and firewall as well as robots.txt, and remember that robots.txt is not a security control.
What about llms.txt?
Treat it as optional documentation. It can help on documentation-heavy sites. There is no validated evidence that it increases citations.
What our own benchmark corrected
In our benchmark of 97 early-stage SaaS websites, 72% had fewer than 100 estimated organic visits a month. Ninety-six percent never appeared on page one for the buyer searches we checked in their own category.
I expected to find many of those sites blocking AI crawlers. Of the 95 we could fetch, 4% blocked any, and 37% already published an llms.txt file. Access was rarely the problem. Being worth citing was.
These figures describe the sites we reviewed. They are not a survey of all SaaS companies, and they don’t show that any one setting causes a traffic outcome.
Promises to distrust
Be sceptical of guaranteed ChatGPT rankings, a fixed number of citations in 30 days, a single “GEO score” with no method, or any claim that llms.txt or schema forces an assistant to cite you. The honest offer is a measurable baseline, a prioritised plan and repeated tracking.
Download now: Free SEO Audit Template [Excel + Google Sheets]
What happens after the audit?
In my experience, the gap between the audit and implementation is where the value disappears.
On one engagement, we re-crawled the client’s site three weeks after delivering the audit and found that none of the recommendations had been implemented. The report was sound. The live site hadn’t changed.
That changed how we work. Earlier in my career I focused on making an audit complete, documenting everything I could find. Now every finding gets an owner, an effort rating and a verification step, and where implementation is in scope, each deliverable can be checked with a crawl.
A few positions I’d defend:
- Judge an audit by what ships. A correct recommendation that never reaches production has no SEO effect.
- Fewer findings are usually better, as long as they are the right ones.
- A health score is a tool’s opinion. A site can score 90 while its most important page isn’t indexed.
- The audit should stand alone. You should be able to hand it to your own developers or another agency.
The step most providers skip is post-fix verification. Releases happen weekly on SaaS sites, and a correct fix can vanish in the next template change.
The growth in our SEO and search-visibility case studies came from programmes that ran for months. An audit identified the constraints. Implementation and iteration produced the results.
Auditing a very large site? Our comparison of enterprise SEO agencies covers who handles audits at scale, and our list of the best AI SEO tools covers what you can check yourself.
Every proposal we send starts with this kind of research; how we work explains what follows, and our pricing page covers the ongoing plans.
Frequently asked questions
What do SEO audit services include?
A real audit covers technical access and indexation, site architecture and internal links, content and search intent, authority and competitor gaps, and AI-search visibility. You should receive prioritised findings with evidence, mapped to URLs or templates, each with an owner and a way to verify the fix.
How much does an SEO audit cost?
A one-off SaaS technical SEO audit typically costs $2,500–$10,000, depending on hosts, templates, rendering, markets and depth. In Ahrefs’ survey of 439 providers, the most common per-project SEO fee was $2,501–$5,000.
How long does an SEO audit take?
It depends on the number of URLs, templates, hosts and markets, and on the access provided. Agree the scope and delivery date at kickoff. Be wary of a deep audit of a complex site promised in 24 hours.
How often should a website be audited?
A stable small site needs a deeper review about once a year, or when the business changes direction. Run targeted audits before and after migrations and after unexplained drops. Monitor indexation, redirects, rankings and conversions continuously in between.
Can I do an SEO audit myself?
Often, yes, for a small stable site with one host, using a crawler and Search Console. The tools show warnings. They don’t tell you which ones matter most, which is where an experienced outside view earns its fee.
The bottom line on SEO audit services
The best SEO audit services are judged by what changes on the live site. Look for evidence, priorities, owners and a verification step, and walk away from scores, exports and guarantees.
If you want to know what is holding your site back, book a search strategy call. We’ll review your visibility on Google and in AI search, then identify the fixes worth doing first. Book a search strategy call through our SEO and AI-search visibility page.
Have an audit already? Bring it for a second opinion. Bring it to a strategy call and we’ll help you separate the findings that matter from the warnings that can wait.
Usman founded Top Search Agency in 2024 and leads strategy on every engagement, from the first AI visibility baseline to the quarterly roadmap, for B2B SaaS and AI companies including Y Combinator-backed Markaz (around 150K monthly organic visits and 444 pages cited in AI answers).
Before that he ran the marketing agency ConvertBunny for more than 20 clients and led digital marketing at nTask, where organic traffic grew from 20K to 280K monthly visitors in 18 months.
He has spent more than a decade in growth and marketing roles, and writes about what actually moves B2B buyers in Google and AI search, based on the programmes his team runs every week.
Connect with Usman on LinkedIn · Meet the team
