The short answerA B2B SaaS CMO should read five SEO KPIs every month: organic-sourced pipeline matched to the CRM, non-brand demo and signup requests, non-brand Google clicks, how often each AI engine names you on fixed buyer questions, and AI-referred visits with what they convert to. Rankings, impressions, CTR and referring domains are diagnostics. They explain why a KPI moved.
| Metric | KPI or diagnostic | What it tells you | Where to read it | How often | What to do when it moves |
|---|---|---|---|---|---|
| Organic-sourced pipeline | KPI | Whether search is putting deals worth money into the pipeline | CRM original source on deals | Read monthly, judge by quarter | Check non-brand requests and clicks before changing the plan |
| Demo and signup requests from non-brand organic search | KPI | Whether people who did not know you convert | CRM original source, split by landing page | Monthly | If clicks hold, fix the path to the form |
| Non-brand Google clicks | KPI | Whether new buyers find you on Google | Search Console Performance report, Non-branded filter | Monthly | Check position and CTR first |
| AI naming rate per engine on fixed buyer questions | KPI | Whether each engine puts you in the answer | An AI visibility tracker such as Peec AI | Monthly | Check whether that engine still reads your pages |
| AI-referred visits and what they convert to | KPI (a floor) | Whether AI answers send buyers who act | GA4 AI Assistant channel, CRM AI Referrals source | Monthly | Compare its conversion rate with organic search |
| Rankings and average position | Diagnostic | Why non-brand clicks moved | Search Console | Weekly | Investigate sudden drops on pages that sell |
| Impressions | Diagnostic | How often Google showed you, AI Overviews included | Search Console, generative AI report | Weekly | Compare impressions with search volume |
| CTR | Diagnostic | Why clicks moved while position held | Search Console, against a current CTR curve | Monthly | Rewrite titles on pages below the curve |
| Core Web Vitals | Diagnostic | Whether slow pages cost conversions | Core Web Vitals report in Search Console | Monthly | Fix failing templates that convert |
| Referring domains | Diagnostic | Why position moved on competitive terms | A backlink tool such as Ahrefs | Monthly | Point new links at the pages that drive requests |
| Indexed pages | Diagnostic | Why impressions dropped on a section of the site | Search Console indexing report | Monthly and after releases | Find what pulled pages out |
| Engagement | Diagnostic | Why traffic converts worse than it used to | GA4 | Monthly | Check the query and page still match |
| Raw AI citation counts | Diagnostic | Which of your URLs AI engines read and quote | AI tracker, retrieved and cited views | Monthly | Strengthen the pages engines read |
What is an SEO KPI, and what is a diagnostic?
A KPI is a number tied to an outcome the business cares about, with a target and an owner who can act on it. The IEEE Technology Navigator draws the line in one sentence: "not every measurable quantity is a KPI, but every KPI must be directly connected to a goal whose achievement or failure matters to the organization." APQC, the benchmarking nonprofit, adds that "There is typically a single KPI backed by supporting indicators used to manage and monitor performance."
Most SEO reports are a wall of supporting indicators with nothing on top. Rankings and impressions went up, and nobody in the room can say whether the pipeline moved. So split every SEO number into two kinds.
A KPI answers a budget question: is organic search, AI search included, putting qualified buyers into the pipeline, and is that getting better or worse? It gets a target and a named owner.
A diagnostic explains why a KPI moved. Average position, impressions, CTR and backlinks live here. They can be urgent, but they don't get a budget target. IEEE warns about "indicator proliferation" and gaming for a reason. Once a diagnostic becomes the target, a team learns to move it without moving revenue.
To test a number, ask whether you would spend more on SEO if it doubled tomorrow and nothing else changed. If yes, it's probably a KPI. If you'd first ask what it did to signups, it's a diagnostic.
What are the five SEO KPIs a B2B SaaS CMO should read every month?
Every search signal gets read against the commercial events on your side: signups, booked demos and any other lead capture you run. A program that moves the visibility numbers and leaves the pipeline flat is a program we call failing, and the report should say so before it shows the chart that went up. This is how we report to clients.
1. Organic-sourced pipeline, matched to the CRM
This is the number the board will ask about, and it lives in the CRM. In HubSpot, the Original Traffic Source property "displays the first known web source through which a contact interacted with your business", and companies and deals associated with that contact mirror it. Organic search and AI Referrals are separate values, which turns organic-sourced pipeline into a report you can defend in a board meeting.
It records first touch only, and HubSpot warns an ad-blocker may block its tracking cookies. Add a "how did you hear about us" field to the demo form. It catches the buyer who met you in a ChatGPT answer and later typed your URL.
Read it monthly and judge it by quarter.
2. Demo and signup requests from non-brand organic search
Pipeline lags. Requests show up first and tell you early whether pipeline will follow.
The non-brand part matters. Someone who searched your company name already knew you, and SEO did not find them. Analytics rarely shows the query behind a conversion, so use a proxy and name it: a request that started on your homepage counts as brand, one that started on a topic page counts as non-brand.
3. Non-brand Google clicks
Search Console defines a click as "The number of times a user clicked your site from Google Search results." Filtered to non-brand queries, it's the earliest KPI that still counts people instead of appearances. Google's own wording for why non-branded queries matter: they "show how new users find your content without any initial intent to go to your site."
Clicks get the target instead of rankings, because the clicks behind a given ranking have shrunk. Pew Research Center tracked the browsing of 900 US adults in March 2025: users who saw an AI summary clicked a traditional result in 8% of visits, against 15% for users who saw none. Ahrefs compared December 2023 with December 2025 across 300,000 keywords and found an AI Overview now correlates with a 58% lower average CTR for the top-ranking page. Seer Interactive then saw CTR on AI Overview queries climb from a floor of 1.3% in December 2025 to 2.4% in February 2026.
None of those is "the" current number. They agree on direction: a ranking can hold steady while the traffic under it drains away.
4 and 5. AI naming rate per engine, and AI-referred visits
KPI 4 is the share of answers to a fixed set of buyer questions that name you, read separately for each engine. KPI 5 is the visits those answers send and what they convert to, read as a floor. Both need their own setup, covered in the AI search section.
Why must brand and non-brand clicks be separate numbers?
Brand clicks measure demand created somewhere else, like a sales conversation or a funding announcement. Fold them into the SEO number and brand growth hides a non-brand decline.
Google made the split much easier. It announced a branded queries filter for the Search Console Performance report on 20 November 2025, and on 11 March 2026 added: "The branded queries filter is now available to all eligible sites." The filter splits clicks, impressions, average position and CTR into Branded and Non-branded views.
Know its limits. Classification is "determined by an internal, AI-assisted system", and Google says "some queries may occasionally be misidentified." It is "only available for top level properties", and only for sites with "a sufficient volume of queries and impressions", which rules out many young sites.
If you can't use the filter and fall back to a regex on the query field, your non-brand total is an undercount. Google omits anonymized queries, those "that aren't issued by more than a few dozen users over a two-to-three month period", from any total filtered by query, and says the match and does-not-match totals "may not add up to the unfiltered total". Write the method next to the number either way.
Brand searches still carry signal if you split them further. New queries pairing your name with a product question mean buyers are researching you. Login and account searches are existing customers going to work, and they inflate "brand growth" every time your user base grows. The Google filter puts both in the Branded view, so keep a short exclusion list of login and account terms to separate them.
How should you read AI search KPIs, engine by engine?
A CMO should care about AI answers because buyers use them. Forrester reported in January 2026 that the share of business buyers using AI in their buying process grew to 94%. G2's 2026 Buyer Behavior Report surveyed more than 1,000 B2B software buyers. Eighty-two percent had sourced software recommendations from an AI chatbot in the past 24 months. The same report found review sites (38%) narrowly overtook AI chatbots (37%) as the top source shaping vendor shortlists.
KPI 4: naming rate, per engine
Peec AI, the tracker we use, defines it plainly: "The Visibility Score measures the percentage of AI responses that mention your brand." Two rules make it a KPI:
- A fixed set of buyer questions. The questions a buyer in your category actually asks, kept the same month to month. Change the set and the number moves for reasons that have nothing to do with your work.
- One number per engine. Never one blended figure.
The second rule comes from our own data. Between June and August 2026, on our own non-branded prompts, our ChatGPT share rose from 5.9% to 17.1% while our Google AI Overviews share fell from 10.1% to 7.1%. The blend moved from 8.3% to 12.3% and showed neither. A blended chart would have reported a clean win while Google's engine slipped.
The engines also disagree with each other. In our study of 987 same-day answer sets, 500 had all three engines naming an agency. In those 500, ChatGPT, Perplexity and Google AI Overviews picked the same agency first 9.6% of the time, and 71% of the agencies named appeared in only one engine's answer.
Our September reading shows how far apart one brand can sit. Over 18 to 28 September 2026, on 141 non-branded prompts: ChatGPT 38.8%, Google AI Overviews 14.3%, Perplexity 11.2%. Each engine needs its own next move.
Share of voice and average position are different numbers, and for a CMO they are diagnostics. Position is "the average ranking across all responses where your brand appears", so it says nothing about how often you show up. The four competing formulas are compared in the AI share of voice formula.
KPI 5: AI-referred visits, and what they convert to
GA4 now has an AI Assistant channel for "sources like ChatGPT, Gemini, Deepseek, Copilot, or Grok". Google's definition adds that the channel "excludes Google's AI Overviews and AI Mode." Those land in Organic Search. A CMO reading "AI traffic" in GA4 is not seeing Google's AI surfaces at all.
In the CRM, HubSpot's Original Traffic Source has an AI Referrals value, with the AI platform domain as the first drill-down, so you can follow an AI-referred visitor into a deal.
Treat every AI referral number as a floor. OpenAI's publisher FAQ says ChatGPT adds a chatgpt.com source parameter to its referral links, and that publishers who allow OAI-SearchBot, the crawler behind ChatGPT search, can track that traffic in analytics. Anything that arrives without a referrer lands in direct, which is why our methodology labels AI-referred visits and AI-attributed leads as floors.
Which diagnostics explain each KPI?
Each diagnostic earns its place by explaining a KPI. If it explains none, drop it from the monthly report.
Average position explains non-brand clicks. Google itself warns that "The position value is a complex metric that can be misleading if you don't understand the subtleties." It averages your topmost position across every query you appeared for, so new loosely related queries can drag it down while clicks grow. For AI Overviews, Google's counting rule is that "An AI Overview occupies a single position in search results, and all links in the AI Overview are assigned that same position."
Impressions explain clicks, and part of AI visibility on Google. For a standard result, an impression counts when your link is on the current page of results, whether or not the user scrolled to it. Inside a carousel or an expanding section, the item typically has to be scrolled or expanded into view. AI Overviews and AI Mode are already inside the normal Web data. Since 3 June 2026, Search Console also has a generative AI performance report, and as of 31 August 2026 it is live for all sites. It shows impressions only, by page, country, device and date. It has no clicks and no queries tab, so you cannot read AI Overview clicks separately in Search Console.
In our experience, some impressions also come from automated queries (AI assistants, rank trackers) that never click. Compare a query's impressions with its search volume before you celebrate impression growth.
CTR explains why clicks moved while position held. Backlinko's often-quoted 27.6% for position one comes from a 2019 dataset that predates AI Overviews. A better yardstick is the organic CTR study from Advanced Web Ranking, built on anonymized Search Console data, refreshed monthly since September 2015 and split by whether an AI Overview appeared. Seer found being cited in an AI Overview delivers 120% more organic clicks per impression than not being cited, yet still 38% fewer than a result with no AI Overview at all on informational queries.
Core Web Vitals explain conversion losses on slow templates. The thresholds, per web.dev: LCP within 2.5 seconds, INP of 200 milliseconds or less, CLS of 0.1 or less, measured at the 75th percentile of page loads. Google adds that good scores don't "guarantee that your pages will rank at the top of Google Search results", and chasing a perfect score "just for SEO reasons may not be the best use of your time."
Referring domains explain position on competitive terms. A rising site-wide total can hide a flat count on the five pages that sell.
Indexed pages explain impressions dropping on a section. A release that adds a noindex tag or breaks a sitemap shows up here before anyone thinks to blame the content.
Engagement explains why organic traffic converts worse than it used to. Usually the query and the page no longer match. Google says clicks from results with AI Overviews "are higher quality", so falling engagement on those pages deserves a close look.
Raw AI citation counts explain the naming rate. Peec separates "Retrieved", where a URL from your domain appeared as a source, from "Cited", where it was cited inline in the answer. Neither is the KPI, but both tell you which pages engines read. In our study of 2,895 AI answers, 98.9% of the answers that named LoudFace had read at least one loudface.co page. The method is in AI engines name the brands they read.
When the numbers disagree, which one wins?
The KPI wins. The mismatch tells you where to look.
AI visibility up, pipeline flat. A visibility rise that never reaches pipeline is a finding rather than a result, and it should change what ships next. Start with which engine rose, since a rise on one engine can sit next to a fall on another. Buyer questions move pipeline far more often than definition questions, so check the question set next. If AI-referred visits did not rise at all, the answers are naming you without sending anyone. If they rose while requests stayed flat, the pages engines read probably lack a clear path to a demo.
Clicks down, pipeline steady. AI Overviews show up mostly on informational searches. Seer found them on 36% of informational queries against 8% of commercial and 5% of transactional ones. Before you treat a click drop as a crisis, check whether the lost clicks were on pages that ever fed a demo.
Impressions up, clicks flat. The generative AI performance report is the first place to check, then automated query volume.
Brand clicks up, non-brand flat. Something outside search is creating demand, and SEO is not the reason. Credit the right team and keep the SEO target on non-brand.
If you are evaluating an agency, ask which of these it reports. A report built only on visibility charts is the warning sign. Our audit of how AI search agencies report pipeline shows what the stronger reports name.
How do you wire the scorecard?
Most of this is configuration you already pay for.
- Search Console on a domain property. Use the Non-branded view for KPI 3 and the generative AI report as a diagnostic.
- GA4 default channel groups. Use the AI Assistant channel for KPI 5, knowing Google's AI surfaces sit in Organic Search.
- Allow OAI-SearchBot. OpenAI says sites that opt out of it "will not be shown in ChatGPT search answers", so blocking it removes you from the answers before you measure anything.
- The CRM. Make sure original source flows from contact to deal, so organic search and AI Referrals reach the pipeline report. Add the self-reported field on the demo form.
- An AI visibility tracker. Load a fixed set of buyer questions and report each engine on its own line. We use Peec AI.
- One page a month. KPIs at the top with targets and owners, diagnostics underneath.
If you want the ROI math and an attribution timeline on top of this scorecard, how to measure AEO agency ROI covers it.
How many SEO KPIs should you track?
Five, and fewer for an early company, where organic-sourced pipeline and non-brand requests may be enough. APQC describes a single KPI backed by supporting indicators. For search, organic-sourced pipeline is that one, and the other four tell you early where it is heading.
Keep as many diagnostics as the team finds useful, as long as none gets a target of its own. Each KPI needs one owner with the authority to act on it. If nobody can change it, it belongs further down the page.
If you want this scorecard run for your company, start with our SEO and AEO service.
