How to measure SaaS SEO in signups and pipeline

Most SaaS teams can say how much organic traffic they get. Fewer can say how many qualified opportunities came from it, or which pages produced them. The data needed to answer that sits in three tools: Search Console knows what people searched, Analytics knows what they did on the site, and the CRM knows whether the account ever bought. Search Console can be linked to GA4, but neither of them knows what happened in the CRM, so the connection to pipeline has to be built.

The SaaS SEO guide sets out the outline of SaaS measurement. The setup below goes into the detail, including the decisions that usually go wrong.

In Google Analytics 4, the actions you care most about are key events: events that measure an action particularly important to the success of the business.¹ I would keep the list short, because a long list of key events makes the reports harder to read, and tie each one to a stage the sales or growth team already recognises.

Leave out newsletter signups, PDF downloads and video plays unless the business treats them as pipeline. They can be tracked as ordinary events without being marked as key.

Analytics knows a session came from organic search. The CRM does not, unless you tell it. The standard way is to write the traffic source and landing page into hidden fields on every signup and demo form, so they are stored on the lead record when it is created.

Which source to store matters more than people expect. GA4 has user-scoped dimensions, prefixed "First user", which record where a new user came from and keep that value when they return. It also has session-scoped dimensions, prefixed "Session", which take a new value each time a user starts a session.² On a SaaS site, a buyer might find a comparison page through organic search in March and come back through a branded search or an email in May to book the demo. G2's 2025 survey of 1,169 B2B software decision-makers found that nearly two in three now prefer to engage with vendor salespeople only in the later stages of their buying journey,³ so the visit that converts is often a long way from the one that started the research. The first-user source credits organic. The session source credits the visit that converted. Both answers are true, and the CRM should hold both. Some CRMs keep a version of each automatically. HubSpot's Original Traffic Source property holds the first known web source and its Latest Traffic Source property the most recent, and HubSpot notes that ad-blockers can block the tracking cookies these values depend on.⁴

GA4's Organic Search channel matches traffic where the source is on Google's list of search sites or the medium is exactly "organic".⁶ GA4 also has a separate AI Assistant channel, which it assigns when the referrer matches its list of AI assistants.⁶ Visits from assistants such as ChatGPT are therefore a different channel from organic search, and your CRM classification should keep them apart too. Other tools draw the same line: HubSpot has an AI Referrals source for platforms such as ChatGPT and Claude,⁴ and Plausible has an AI Assistants channel.⁷ Visits that arrive with no referrer cannot be identified as coming from an assistant at all. Browsers send no referrer when a page or link uses the no-referrer policy,⁸ and clicks from emails, documents, messaging apps and mobile apps often arrive without one.⁷ Seer Interactive's testing across client accounts, updated in June 2026, found that ChatGPT usually adds UTM parameters to links it shows as citations, and that visits without them or a referrer are likely to be reported as Direct in GA4.⁹

For most B2B SaaS companies, organic pipeline per URL is a handful of opportunities spread across hundreds of pages. That is too thin to read. Group the landing pages by type, using URL patterns where the site structure allows, and report every metric by group.

The expectations in that table are typical patterns, not benchmarks. Your own split is the useful number. If the comparison pages produce more opportunities than the whole blog from a fraction of the traffic, that is a reason to move effort, and the quality of those pages starts to matter a great deal. I have written up how to build comparison pages that buyers trust.

Once the lead record carries its organic source and landing page group, the CRM can report opportunities and revenue by those fields. That is the report that answers whether SEO is producing pipeline. It will be small and it will lag by the length of the sales cycle, so a quarter's organic signups may not show their revenue for several quarters.

Some teams also want CRM stages back in Analytics, so they can see opportunities against landing pages in GA4 reports. GA4's Measurement Protocol lets you send events from a server, and its documented uses include sending offline conversions and tying online to offline behaviour.¹⁰ It uses the client ID to join those events to the web activity GA4 already holds,¹⁰ which is why storing the client ID on the lead record is worth doing even if you do not use it straight away. The protocol is meant to augment tagging, not replace it.¹⁰

Branded and non-branded search measure different things, and reporting them together lets a successful launch or a busy events season make SEO look better than it is. Search Console's branded queries filter splits Performance report data into branded and non-branded views. The classification is made by an AI-assisted system that includes brand name variants, typos and queries about products unique to the site, and Google's announcement admits some queries may occasionally be misidentified.¹² It is only available for top-level properties with enough query volume, not for URL-path or subdomain properties.¹² The Insights report has a matching card that shows branded against non-branded clicks.¹²

The CRM cannot make that split directly, because the query is not passed to the site with the visit. A workable proxy is the landing page group. Organic leads that enter on the homepage, pricing or login pages are mostly branded. Leads that enter on comparison, integration, use-case or blog pages are mostly non-branded. It is imperfect, and I would say so in the report, but it gives a pipeline view of the two kinds of demand that is good enough to act on.

Search Console is the only source of query and impression data from Google Search, and that makes it essential. It also has limits that matter more on a niche B2B site than on a large consumer one.

• Queries that are not issued by more than a few dozen users over two to three months are anonymised and left out of the query rows, though their clicks are counted in the chart totals.¹³ Specialist software queries can easily fall into that group.
• Anonymised queries are omitted whenever a filter is applied, so the totals for queries that contain a term and queries that do not will not sum to the overall total.¹³
• The interface exports at most 1,000 rows, and the Search Analytics API returns up to 50,000 rows per day per site per search type, according to Google's 2022 post on these limits.¹³
• The bulk data export to BigQuery includes all the performance data available for the property, except anonymised queries.¹⁴ It is the better option once the site has more pages and queries than the row limits allow.
• It reports nothing about what happened after the click. Signups, opportunities and revenue have to come from Analytics and the CRM.

Use Search Console for what it is good at: which queries and pages gain or lose impressions and clicks, and how branded and non-branded visibility move over time. Use the CRM for whether any of it turned into revenue.

• Non-branded clicks and impressions from Search Console, by landing page group, against the same quarter last year.
• Organic key events from GA4, by landing page group, with signups and demo requests shown separately.
• Organic leads, opportunities and revenue from the CRM, split by first-touch and converting-session source.
• The same CRM figures split by the branded and non-branded landing page proxy, with the method stated.
• Notes on anything that distorts the quarter: a launch, a pricing change, a tracking fault, or a change in GA4 settings.

A report like this will show smaller numbers than a traffic chart. They are the numbers a finance director will believe, and they tell you which page types to build more of.