B2B buyers are 61% of the way through their buying journey before they first contact a vendor, and the winning vendor was already on their day-one shortlist 95% of the time, according to 6sense's 2025 Buyer Experience Report of nearly 4,000 buyers. Most of that first 61% happens on websites, yours included, and almost none of it shows up as a form fill. That is why a B2B website dashboard built on traffic totals tells a team so little.
This guide lists the 11 metrics we think deserve a place on the dashboard, the Google Analytics 4 (GA4) event name or Search Console report behind each one, and six numbers that look like progress but rarely change a decision. Definitions come from Google's own Analytics, Search Central and web.dev documentation, read in October 2026.
What makes a metric worth a dashboard
A metric earns a place on the dashboard if a change in it would make someone act differently on Monday. Rising traffic fails that test: nobody knows whether to write more, spend more or fix something. A falling form start rate on the demo page passes it, because it points at a page, a form and a likely cause.
So each metric needs a funnel stage, a named owner and a definition that does not change month to month. If you are still choosing the tools, our guide to the analytics stack for a four-week site build covers GA4 limits, consent mode and heatmaps; this page decides what to read from that stack.
The 11 metrics, with the GA4 event behind each
The list runs in funnel order, from visibility to pipeline, then site health. Each row names the event or report that feeds it.
| # | Metric | Source event or report | Formula | Review |
|---|---|---|---|---|
| 1 | Non-branded organic clicks | Search Console Performance, Web search type | Clicks on queries without the brand name | Monthly |
| 2 | Engagement rate by channel | GA4 engaged sessions | Engaged sessions / sessions, per channel | Monthly |
| 3 | High-intent page reach | GA4 page_view on pricing, case studies, integrations | Sessions with one of those views / sessions | Monthly |
| 4 | Form start rate | GA4 form_start | Sessions with form_start / sessions that viewed the form page | Weekly |
| 5 | Form completion rate | GA4 generate_lead and form_start | generate_lead / form_start | Weekly |
| 6 | Visitor-to-lead rate by channel | GA4 session key event rate on generate_lead | Sessions with generate_lead / sessions | Monthly |
| 7 | Leads by form type | generate_lead with lead_source and a form type parameter | Count per form type | Weekly |
| 8 | Lead-to-qualified rate | qualify_lead and disqualify_lead from the CRM | qualify_lead / generate_lead | Monthly |
| 9 | Website-sourced pipeline value | qualify_lead or close_convert_lead with value and currency | Sum of value, by first channel | Monthly |
| 10 | Self-reported source | "How did you hear about us" field on the form | Share of leads per answer | Quarterly |
| 11 | Core Web Vitals pass rate | Search Console and field data at the 75th percentile | Templates passing all three / templates | Monthly |
Source: Google Analytics Help, Google for Developers recommended events, Google Search Central, web.dev; formulas are ours
Non-branded organic clicks. Non-branded clicks measure whether people who did not know you are finding the site. Google Search Central says traffic from AI Overviews and AI Mode is reported inside the same Performance report under the Web search type, with no separate line, so falling non-branded clicks on stable impressions is the first sign that answers are being given above your result.
Engagement rate by channel. Engagement rate is the share of engaged sessions, and it is worth reading only per channel. A site-wide figure mixes paid social visitors who leave at once with organic readers who stay.
High-intent page reach. Pricing, case studies, security and integration pages are where shortlisted buyers check you. The share of sessions reaching one of them is the best pre-form intent signal GA4 gives you without extra tools.
Form start and completion. These two together locate the leak. A low start rate is a page problem: offer, copy or trust. A low completion rate is a form problem: field count, errors or a calendar step. Our post on B2B website conversion rate walks through finding that leak page by page.
Leads by form type. A demo request, a content download and a newsletter sign-up are not the same lead. Send generate_lead for all of them, with a parameter that says which form fired it, and keep the demo and contact forms as the only ones that are key events.
Pipeline metrics. Metrics 8 and 9 are the ones sales will read. They need the CRM to send qualify_lead, disqualify_lead and close_convert_lead back into GA4 with a value and currency, which Google's recommended events reference describes as typically required for meaningful reporting.
Self-reported source. Clicks cannot see a podcast, a peer recommendation or a Slack community. A free-text or short-list "How did you hear about us" field on the demo form is crude, and it is still the only measure of the 61% of the journey that happens before contact. We found no independent study with a reliable share for these channels, so read your own answers rather than a benchmark.
Core Web Vitals pass rate. Count templates, not pages: home, solution, case study, blog post, pricing. A template passes when LCP, INP and CLS all meet the good threshold at the 75th percentile, separately for mobile and desktop.
An event naming plan you can hand to a developer
GA4 has three kinds of event: enhanced measurement events fire automatically once switched on, recommended events have names and parameters Google has already defined, and custom events are everything else. Prefer them in that order.
| Event | Kind | Fires when | Key event | Parameters to send |
|---|---|---|---|---|
| page_view | Enhanced measurement | Page loads or browser history changes | No | Default |
| scroll | Enhanced measurement | First time a user reaches 90% depth | No | Default |
| file_download | Enhanced measurement | Click on a link to a PDF, DOCX, ZIP or similar file | No | Default |
| form_start | Enhanced measurement | First interaction with a form in a session | No | Default |
| form_submit | Enhanced measurement | A form is submitted | No | Default |
| generate_lead | Recommended | The form's success state is shown | Yes, for demo and contact forms | lead_source, currency, value, form type |
| qualify_lead / disqualify_lead | Recommended | Sales marks the lead in the CRM | qualify_lead yes | currency, value, disqualified_lead_reason |
| working_lead | Recommended | A representative and the lead make contact | No | currency, value, lead_status |
| close_convert_lead / close_unconvert_lead | Recommended | The deal is won or lost | close_convert_lead yes | currency, value, unconvert_lead_reason |
Source: Google Analytics Help, Enhanced measurement events; Google for Developers, Recommended events (updated 2026-09-16)
Two details catch most teams. First, fire generate_lead on the success state, not on the click: form_submit only says a form was submitted, while generate_lead should mean a lead your CRM actually received. Second, the parameter is called disqualified_lead_reason, not disqualified_reason. A misspelt parameter does not error; it just never appears in a report.
For a first report, Google's help article on lead generation forms builds a funnel exploration from page_view, form_start and form_submit. Explorations can take up to 24 hours to show new events.
The 6 metrics that are vanity on a B2B site
None of these is useless. Each one is vanity when it sits on a dashboard as a headline number, because it can rise while revenue falls.
| Vanity metric | Why it misleads on a B2B site | Read instead |
|---|---|---|
| Total sessions and users | One post ranking for a broad term can lift traffic with no buyers in it | Non-branded clicks and high-intent page reach |
| Site-wide bounce rate | In GA4 it is the share of sessions not engaged; one 11-second visit counts as engaged | Engagement rate per channel and per template |
| Average engagement time | Averages are pulled up by a few long sessions and say nothing about intent | Form start rate on the pages that matter |
| Views per session | GA4 counts repeated views of the same page, so a confusing navigation raises it | High-intent page reach |
| Site-wide average position | Search Console averages your topmost result across every query, branded included | Non-branded clicks and impressions per page |
| Total key events | Once scrolls or downloads are marked as key events, the total grows with traffic, not leads | generate_lead by form type, then qualify_lead |
Source: Google Analytics Help, Engagement rate and bounce rate; Analytics dimensions and metrics; Search Console Help, Performance report
Bounce rate survived the move to GA4 with a different meaning. Google now defines an engaged session as one that lasts longer than 10 seconds, has a key event, or has two or more page or screen views, and bounce rate as the share of sessions that were not engaged. A buyer who reads your pricing page for nine seconds, takes a screenshot and sends it to their finance lead is a bounce. A bot that sits on the page for 11 seconds is not.
Benchmarks: use them for the shape, not the target
Published benchmarks are useful for spotting a broken stage, less useful as a goal. The most detailed public funnel data we found comes from First Page Sage, an SEO agency reporting on more than 50 B2B SaaS clients, mostly with $10 million to $100 million in revenue. Treat it as one agency's client base, not a market average.
| Channel | Visitor to lead | Compared with SEO |
|---|---|---|
| 2.2% | Slightly higher | |
| SEO | 2.1% | Baseline |
| 1.3% | About 38% lower | |
| Webinar | 0.9% | About 57% lower |
| PPC | 0.7% | About two thirds lower |
Source: First Page Sage, B2B SaaS Funnel Conversion Benchmarks, updated 2025-06-11 (agency client data)
The spread matters more than any single figure. Paid search converts visitors at a third of the rate of SEO in this data, so a month in which paid spend rises will lower the blended visitor-to-lead rate even if every page improved. This is the reason metric 6 is read per channel. For form-level benchmarks, our page on demo request conversion rates collects the published medians.
Who owns which number, and how often to look
Marketing owns 1 to 3. These respond to content, positioning and channel spend. Review monthly; weekly movements are mostly noise on a low-traffic B2B site.
The web team owns 4, 5 and 11. These respond to page and form changes, so review them weekly and after every release. A release that hurts form completion should be rolled back within days, not found at quarter end.
Marketing and sales share 6 to 10. These only mean something when both teams agree what a qualified lead is, so write the qualify_lead definition down before building the report. If a metric has not changed a decision in two quarters, move it to a detail report.
How to set this up on a new or rebuilt site
Measurement is cheapest to get right during the build, while forms and templates are being decided anyway. In our four-week marketing-site process we would agree the event list in week one and test every key event on staging before launch.
Week one: write the plan. List the forms, high-intent pages and CRM stages, and map each to an event name.
During the build: fire generate_lead on success. Add the form type and lead_source parameters, and switch on enhanced measurement for scroll, file downloads and form interactions.
Before launch: mark key events. Mark generate_lead (filtered to demo and contact forms where your setup allows), qualify_lead and close_convert_lead, and nothing else.
After launch: connect the CRM. Send qualify_lead, disqualify_lead and close_convert_lead back with value and currency. Until this is live, metrics 8 and 9 stay empty, and the dashboard should say so rather than show zeros.
What this means for your B2B dashboard
Most B2B dashboards carry too many numbers from the top of the funnel and none from the bottom. Cut the six vanity metrics from the headline view, build the 11 above in the order given, and accept that the pipeline rows will take a few weeks of CRM work before they fill in. The strategic choices that sit upstream of measurement, such as which pages the site needs and who it is for, are covered in our B2B website strategy guide. If you are rebuilding the site anyway, our marketing websites service page explains how we scope measurement into the build.
