SaaS SEO Audit: What Your Consultant Should Diagnose and Deliver

A SaaS SEO audit is a review of how a software company's website turns search demand into qualified demos, trials and enquiries. It checks whether the right pages are visible for commercially valuable searches, whether searchers click, whether the page moves them to a next step, and whether the leads that result are worth having. Technical health is part of it, the part that lets the site be found and understood in the first place.

What makes it a SaaS audit rather than a general website audit is the shape of the path. A SaaS site carries product, use-case, comparison and integration pages that have to win buying searches, and often a blog, docs or a help centre that attracts far more traffic than the pages that sell. The conversion is rarely a purchase. It is a demo request that sales will accept or decline, or a trial the product will activate or lose. And the person searching is often one member of a buying group, so the page that ranks may have to convince someone who will not be the one signing.

That changes what "working" means. A page can rank, draw clicks and convert at a healthy rate while sending sales a steady flow of people who were never going to buy. A technical audit will score that page well. A SaaS SEO audit should flag it.

You need an audit when there is a commercial question that the team cannot answer from normal reporting.

Typical moments include:

• Organic clicks, rankings or qualified leads are falling, and nobody knows whether the cause is demand, visibility, the results page, the website or tracking.
• The site gets organic traffic but few demos, trials or useful enquiries.
• Sales says organic leads are weak, but marketing cannot see which pages or search clusters produce them.
• A content library has grown without a clear page or cluster strategy, and pages now compete with each other.
• The company is about to redesign, migrate, merge websites or change its CMS and needs to protect the search footprint that already works.
• Paid search is carrying most pipeline and the team needs to understand where organic search could create a more durable path.
• A new market, product, positioning or buyer segment changes which demand and pages matter.

How often should you run one?

Not every audit is a recurring project. There are three kinds, and each asks a different question.

• Continuous monitoring

  Monitoring checks that nothing has broken. It means light technical and measurement checks each month, especially after releases. On large sites with thousands of pages, monthly technical checks per domain are sensible, because at that scale templates, releases and content changes can break something most months. Smaller sites need lighter checks, mainly on the settings that go wrong quietly: canonicals pointing at the wrong page, noindex left on after a release, redirects that chain. This is routine work with a crawler and Search Console alerts. It should not be sold as a project. It is part of ongoing search growth and maintenance.

• Milestone audit

  A milestone audit happens before a redesign, migration, rebrand, website merge or major template change. Its job is to protect what already works: which URLs carry the rankings and the leads, which content should be moved, merged or retired, and which redirects cannot be missed. I have worked through migrations with new templates, redirect mapping and a content purge across thousands of pages, and the audit that mattered most was the one done before anything moved. It is also the moment to agree with the UX and design team how the new templates are structured, so the rebuilt site supports search from launch rather than being retrofitted afterwards.

• Diagnostic audit

  A diagnostic audit starts when commercial search performance changes, conversion quality is unclear, or a team needs to decide what to fix before investing further. The rest of this article is mostly about this kind of audit.

A full commercial audit is not something to schedule automatically every quarter. Run it when a meaningful decision depends on it. Between those moments, monitor the agreed priority pages and search clusters so problems are found before they become a large audit.

A useful audit looks at the whole path from search to customer. Seven areas cover it, and each answers a different part of the commercial question. They overlap, and the value comes from reading them together rather than scoring each one on its own.

Demand and search performance

Start with what search already produces. Search Console shows impressions, clicks and positions by query and page. Grouping them into clusters shows where you are visible for demand that matters commercially, and where you are not. Set that against the live results for your priority searches: what actually ranks, which page types win, and whether competing is realistic.

This is also where changes in demand itself show up. When a whole topic loses impressions while your positions hold, the problem may not be on your website at all.

A SaaS SEO audit should map the current keyword universe, not only review isolated pages. Group existing ranking queries, priority non-ranking queries and pages into demand clusters and buying stages: problem-aware, solution-aware, category, use case, comparison, integration and product evaluation.

For each cluster, establish:

• the buyer question and stage
• the queries already producing impressions or clicks
• the page meant to win the cluster
• the pages actually competing or ranking
• the intended next action
• whether the cluster is missing a page, has the wrong page type, or has too many pages competing

Content effectiveness and commercial-page coverage

Check whether a page exists for each priority cluster, whether it is the page type the results reward, and whether it is the page that actually ranks. SaaS sites often have the product, use-case, comparison and integration pages a buyer needs, while the traffic lands on a blog post or a help article instead. A help centre can hold more of the search footprint than the marketing site, and it is often left out of audits because a different team owns it.

Then look at the content that does not earn its place. A mature SaaS blog usually contains posts competing for the same query, posts that decayed years ago, and posts that attract readers nobody would sell to. An audit does not default to publishing more content. For each cluster, it should identify whether the best move is to:

• strengthen an existing page
• merge overlapping pages
• redirect or retire thin or obsolete content
• improve internal links and the route to a commercial page
• create a missing page, only where there is a genuine demand and coverage gap

A smaller, sharper search footprint often competes better than a large, diluted one.

Content cannibalisation, consolidation and page strength

When several pages target the same buyer question, they can send mixed signals to search engines about which one should rank. They also split the links, clicks and conversion evidence that one stronger page would otherwise collect, which makes every one of them harder to judge. The audit should decide which page becomes the primary page for the cluster, which pages should support it, and which should be merged, redirected or removed.

A content purge is not about reducing page count. It is about making the remaining search footprint clearer, stronger and more commercially useful.

Website structure and technical health

This is the backbone. If search engines cannot crawl, render and index the priority pages, nothing else in the audit can pay off. The technical work stays in, focused on the pages that serve priority clusters rather than the whole crawl. On SaaS sites, the issues I come across most often include:

• JavaScript-rendered content that search engines only partly see
• docs and help-centre pages duplicating marketing pages, or versioned docs competing with each other
• canonicals and noindex set at template level and applied to the wrong page types
• redirect chains left over from earlier migrations
• hreflang errors where the product is sold in several markets

Structure matters as much as individual errors. The navigation and internal links tell search engines which pages the business considers important, and in my experience they often point at the blog rather than at the product and comparison pages that sell.

AI visibility belongs here as a supporting check rather than a separate workstream. Whether AI answers describe your product accurately for your priority searches, and whether your pages are easy for them to use, depends on the same things as organic visibility: pages that answer buyer questions with proof, on a site search engines can read. I set out why in the Strategy article.

B2B messaging, proof, UX and conversion

Once someone arrives, the page has to do commercial work. The audit checks whether it matches what the searcher wanted, whether it carries proof the people evaluating the product find credible, and whether the layout makes that easy to find on mobile as well as desktop. It also checks the next step: whether the call to action suits where the visitor is, and whether the form adds friction a qualified buyer would abandon over. A demo request is a large ask from someone still comparing options, and a page that offers nothing smaller can lose them.

This is where SEO and conversion work meet. In my experience it is also where a lot of organic value leaks without ever showing up in an SEO report.

CRM, lead quality and product activation

The conversion is not the finish line. Where the product is demo-led, the audit should look at whether sales accepts the demos that come from organic search, and which clusters produce opportunities. Where it is trial-led, it should look at whether organic trials activate. Sales notes on recent organic leads add the reasons behind the numbers.

This evidence tells you whether a page that converts well is sending the right people. Without it, the audit can only judge the path up to the form.

Measurement and implementation feasibility

Two things decide whether the findings can be trusted and acted on. The first is measurement. If key events, lead-source fields or consent settings are wrong, every other finding rests on bad numbers, so the audit should check the tracking before it trusts the dashboards.

The second is feasibility. SEO fixes rarely ship on their own. A new comparison page goes through the brand team, a template change joins the web product backlog, and a change of message needs product marketing to agree. In-house, I handed findings to the content team, to developers and to the brand managers who owned sections of the site. The findings that got done were the ones written for their roadmaps, not for mine.

Where a redesign, migration or site merger is planned, feasibility also means risk. The audit should say which URLs, pages and redirects carry the rankings and leads that must survive the change.

The same symptom, fewer demos from organic search, can come from six different constraints anywhere on the path from search to customer. Each needs different people and different work. Fixing the wrong one can cost a quarter and make the real one harder to see.

The constraint locator below is how I separate them. Each row is a place where the path can break, from the demand itself through to the numbers you use to judge it. The columns show what you would see, how to confirm it, what it is usually mistaken for, and who tends to own the fix.

Three things matter when you use it.

Work from the start of the path. If the demand has shrunk or was never the right demand, no page fix will recover it. If priority clusters are not visible, a better demo page will not help much, because the right people are not reaching it. If sales declines most organic demos, more traffic makes the problem larger. The earliest broken stage on a valuable path is usually the one to confirm first.

Confirm before you conclude. Every row has a likely misreading, and the evidence under "How to confirm it" is what tells them apart. A falling click-through rate may point at the snippet, or at a results page that has changed around you. Until you have looked at the live results for those queries, you cannot tell which. Technical findings need the same test. A crawl will always find something, and that is not the same as finding the constraint.

Rule out measurement early. It appears last in the table because it can imitate any other constraint. A consent banner change, a key event redefined in analytics or a CRM rule that overwrites lead source can move every number on a dashboard without a single buyer behaving differently. It is the cheapest constraint to rule out, and the most frustrating to discover after a redesign.

When clicks and impressions are both falling

This is the most common reason teams ask me for an audit right now, and in my experience it is rarely a penalty. The loss usually sits in one of four places: rankings lost on specific clusters, a results page that now answers the query before anyone clicks, lower demand for the topic itself, or a change in tracking. Search Console by query and page, set against the live results and a record of what the team released, will usually separate them. Where the decline is concentrated in question-shaped informational queries while commercial queries hold, the changing results page is a likely cause. Where commercial queries are falling too, the site itself deserves a closer look before anyone blames AI summaries.

Scope is usually set by page count: the whole domain, every URL, every issue. That produces a thorough audit and a long list, and the list tends to be ordered by whatever is easiest to count.

I would rather set scope by demand. Start from the three to five search clusters that matter most commercially, then include the pages that serve them, wherever they sit. Sometimes that is the whole site. Often it is narrower: only the industry pages, only the comparison pages, or only the help centre, because that is where buyers evaluating the product actually land.

Within that scope, an audit that finds forty real problems still has to say which three matter. I rank findings on four things.

• Commercial value of the cluster

  Value is how much qualified pipeline or activation that demand could plausibly produce, judged from CRM and product data where you have it, not from search volume.

• Confidence

  Confidence is how sure the evidence makes you that this is the constraint. A finding confirmed in Search Console, analytics and sales feedback outranks one that rests on a crawl alone.

• Effort

  Effort is the work involved across content, design, development and review.

• Ownership

  Ownership is whether a named team can take it on this quarter. A high-value fix with no owner is a decision to make, not a task to assign.

Two kinds of finding often rise to the top. The first is a commercial page that almost works: visible, clicked, and losing people at a fixable point, such as a missing comparison, thin proof or a next step that asks too much too early. The second is a small cluster that produces unusually good conversations. It is easy to overlook because the volume is low, and it is often the best candidate to protect and expand.

At the bottom sit findings that are real but cannot change the commercial result, such as a warning on a page nobody should land on, or a rule broken on a template that carries no priority demand. They belong in the monitoring backlog, not on the audit's first page.

A crawl tells you how the site is built. It does not tell you what search produces. The audit needs Search Console, analytics, the CRM, product analytics where the product is trial-led, and a review of the live results for priority searches. I covered what each source shows in the Strategy article. In an audit, the harder work is joining them.

Most SaaS teams I talk to have the data and cannot connect it. The gaps I come across most often include:

• key events in analytics counting page views or button clicks rather than completed demo requests or signups
• lead source in the CRM overwritten by the last touch, so organic demos appear as direct or email
• no link between the page someone landed on and their lead record
• organic trials that cannot be separated from other trials in product analytics
• no record of when templates, redirects or tracking changed

Some of these become findings in their own right. If sales acceptance by landing page cannot be measured, the audit should say so, recommend the fix and use what is available in the meantime, such as sales notes on a sample of recent organic leads or a self-reported source field. That is less precise, and still far better than judging lead quality from form fills.

One source is easy to forget: the release log. Many ranking and conversion changes line up with a deployment, and nobody connects the two without it.

For access, the audit usually needs Search Console, analytics, permission to crawl the site, CRM reporting by lead source and landing page, product analytics for trial-led products, and half an hour with someone in sales.

You do not need an agency to run a useful first pass. You need a narrow scope, access to the evidence above, and the discipline to stop at the first real break rather than listing everything. These are the six steps I would follow.

1. Choose three to five commercially important search clusters

   Pick them by the buyers and deals they could produce, not by search volume. A comparison cluster with modest demand can matter more than a large informational topic. Write each one down as a buying moment, such as teams comparing you with the tool they already pay for.

2. Map the pages and the intended next action

   For each cluster, list the page meant to win it, the pages that actually rank or receive the traffic, and the action you want the visitor to take next: a demo, a trial, a pricing visit or further reading. Gaps and mismatches often show up here, before any data is pulled.

3. Map the keyword universe by cluster and stage

   Group the queries and pages from Search Console by cluster and buying stage. For each cluster, flag one primary page, any overlap or cannibalisation, content to consolidate, and genuine coverage gaps.

4. Join the evidence by cluster

   Group Search Console queries and landing pages by cluster, then follow the same pages into analytics, the CRM and product data. Where the join breaks, note it. That gap is often a finding in its own right.

5. Confirm the earliest meaningful break

   Use the constraint locator to find the first stage where the path fails for each cluster, and check the evidence against the likely misreading before you conclude. Rule out measurement early.

6. Write findings in one format

   Record what you saw, the evidence, how confident you are, the likely owner and the effort involved. Findings in the same shape can be compared, and the weak ones become obvious.

Whoever runs it, in-house or outside, a useful audit should hand over:

• a map of priority search clusters against the pages meant to win them
• findings placed on the path, each with its evidence and how confident that evidence makes you
• a short, ordered list of what to do first, with the effort and a likely owner for each item
• a check of each priority fix against the brand, content and product roadmaps already in motion
• a walkthrough with the people who will do the work, not only the person who commissioned it

It should not be a spreadsheet of every issue a tool found. Tool output is an input. It becomes an audit when someone decides which findings matter commercially, and in what order.

How to judge an SEO audit or agency

Ask how findings are ranked. If the answer is severity or issue count, the audit will be accurate and hard to act on. Ask which data they will use beyond the crawl, and whether CRM or product data is part of it. Ask who implements, and how the findings will fit your existing roadmaps. Then ask what the audit would conclude if the real constraint turned out to be the offer or the sales process rather than search. A good auditor will tell you, even when the answer sits outside their service.

An audit on its own is enough when the demand is right, the constraint is clear and your teams have the capacity to fix it. That happens more often than you might expect. A focused audit often ends with a handful of changes an in-house team can ship without outside help.

It is not enough in three situations.

• When the demand itself is wrong

  If the site is visible, clicked and converting for searches that produce the wrong buyers, a better page will not fix it. The next step is a decision about which demand to pursue, which is strategy work.

• When findings compete for the same teams

  If the priority fixes all need the same developers or the same brand approvals, you need a roadmap that sequences them against everything else those teams are already doing.

• When nobody can implement

  If the fixes cross content, design, development and product, or involve new templates, someone has to carry them through. An audit that nobody implements is a well-evidenced opinion.

Can an SEO audit increase demo requests and trial sign-ups?

It can show where they are being lost and what is most worth changing to recover them. The audit alone changes nothing on the site. What it can do is stop a team spending a quarter on work that could not have moved the result, and point that effort at the constraint that can.

Prices vary widely, because "SEO audit" covers very different pieces of work. An automated tool report, a checklist technical audit and a commercial diagnosis of the path from search to customer are all sold under the same name. They cost different amounts because they need different amounts of senior judgement and time with your data.

The main things that drive the cost of a commercial audit are:

• the number of templates and page types, more than the number of URLs
• how many priority clusters are in scope
• the number of domains, languages and markets
• whether CRM and product data are available, and how much work joining them takes
• whether it is a milestone audit, where redirect mapping and migration planning add work
• whether it ends in a prioritised roadmap and a live walkthrough, or stops at findings

Whatever the price, expect an agreed scope and data access, findings tied to evidence, a prioritised sequence with owners, and a session with the people who will act on it. If a proposal lists tools and page counts but not how findings will be ranked, you are probably buying a crawl report.

At GrowthLens, the scope and fee are fixed and agreed after an Opportunity Review, once I know which clusters and which part of the path the audit needs to cover.

A successful audit should not end with a long list of issues. It should pinpoint at least five valuable actions the team can take now or sequence into the next 90 days.

For every priority action, the audit should state:

• the demand cluster and page affected
• the constraint it addresses
• the proposed action
• the expected measurement signal
• the owner and effort
• what to measure before the change, and again after it is live

That makes the audit useful as a baseline as well as a diagnosis. Once the work is implemented, the team can measure whether visibility, clicks, conversion, sales acceptance or activation improved. None of these actions guarantees revenue. What the audit gives each one is a measurable hypothesis, and a before-and-after basis for judging whether the change worked.

If a search cluster should be producing more pipeline and you cannot see why it is not, request an Opportunity Review. I will look at one priority path with you, tell you where I think the constraint sits, and whether an audit is the right next step.