Pageview Analytics: Traffic Up, Signups Flat
Pageview analytics is good at one question. How many people loaded the page, and from where. I still want that number. It is the wrong tool for the next question, which is where the signup path died.
I hit this on a week when the homepage and /signup were both up and new accounts were not. The chart looked fine. Support did not. Someone had started the form and never created a project, and pageview analytics could not point at that step. I could see the visit. I could not see the drop. A conversion funnel could, once the steps had names.
The setup for those steps is cookieless conversion funnels. Stay on this page if you are about to throw out a traffic tool for a question it was never going to answer.
The storage claim, that the script sets no cookie, is cookieless analytics. The longer checklist, including what I will not promise, is privacy first analytics.
Two questions, two tools
"How many people visited?" is a traffic question.
"Where did the flow break?" is a product question.
Cookieless pageview analytics tools, Plausible and Fathom among them, answer the first one cleanly. Fast script, no cookie banner maze. That is the floor. The search that usually follows is a Plausible alternative or a Fathom alternative, because traffic went up and signups did not (EventDash vs Plausible, vs Fathom).
The dashboard can look healthy while the business is stuck. Pageviews do not name the step that failed. That is a category difference, not a bug in the traffic tool.
What a visit chart hides
A typical onboarding path I watch:
- Land on the homepage
- Click "Start free"
- Finish the signup form
- Create the first project
- Invite a teammate
Pageview analytics will show homepage visits and /signup visits. It may show time on page. It will not show that people who started signup never created a project, or that invites dropped after a pricing change. I do not get a percentage from the visit chart. I get a shrug.
A conversion funnel shows the drop between those steps. One look, once a week, one product fix. That is funnels and retention, and it is not bounce-rate theater.
Signs the questions outgrew the chart
I treat these as the switch:
- Support says "I signed up and got stuck" and you cannot see where
- Marketing celebrates traffic while product cannot explain activation
- You export a CSV to guess the steps
- Roadmap arguments end with "we need more data" and nobody owns event names
None of that means the traffic tool failed. It means the question changed. EventDash is the layer I add for funnels, HTML goals, and events. Naming those events is event tracking basics.
What I add first
I do not instrument fifty events on day one.
Week 1: one activation funnel, three to five steps, and two goals on real buttons (signup_clicked, upgrade_clicked).
Week 2: open the funnel on Monday, fix the worst step, write down what shipped.
That is enough for a small team. Expand when a new workflow becomes the question.
Marking the goals without a tag manager is track conversions without Google Tag Manager.
What I still skip
Product analytics does not have to mean a platform.
On day one I do not need session replay, feature flags, a data team, or a Mixpanel-sized cohort board. PostHog ships replay and flags. That is a product OS. EventDash is the PostHog alternative when I wanted the analytics and not the rest of the surfaces. A pageview tool still answers visits. It does not answer drop-off.
Pick the question
| Weekly question | What I use | | --- | --- | | How much traffic, from where? | Pageview analytics | | Where does signup die? | EventDash: funnels and HTML goals | | Replay, flags, experiments | A platform you will actually operate |
Some teams run both for a while. Fine. What fails is one visit chart pretending to answer both.
What I do when the charts disagree
- Ignore bounce rate for a week
- Open one activation funnel
- Find the largest step drop
- Ship one change aimed at that step
- Check again next Monday
Free is 100k events/month. Signup if the missing piece is the funnel, not another traffic chart.
Sources
Google's page-view reporting model is documented in GA4 views. What a conversion funnel is, in one sentence, stays on the glossary.
Related
Key takeaways
- Pageview analytics answers who arrived. Product analytics answers where they stopped.
- Traffic up and signups flat is a funnel problem, not a traffic problem.
- Start with one activation funnel of three to five steps and two CTA goals.
- You still do not need session replay, feature flags, or a data team on day one.
FAQ
- What does pageview analytics actually tell you?
- Pageview analytics tells you who arrived and which URL they loaded. It is the right tool for traffic, referrers, and top pages. It does not tell you which product step died after the click. When traffic goes up and signups stay flat, I stop staring at the visit chart and open one activation funnel instead.
- When should I add a funnel on top of pageviews?
- When support says someone signed up and got stuck, and you cannot see the step. When marketing is happy about traffic and product cannot explain activation. When you are exporting a CSV to guess the path. That is the moment pageview analytics has done its job and the next question is a different one.
- What should I measure first?
- One activation funnel with three to five steps, and two goals on real buttons, such as signup_clicked and upgrade_clicked. I open it on Monday, fix the worst step, and write down what I shipped. Fifty events on day one is how the model gets noisy before it gets useful.
- Do I have to throw away the pageview tool?
- No. Some teams keep a pageview tool next to EventDash. That is fine. What fails is pretending the visit chart answers drop-off. Replay, feature flags, and a data team can wait. I wanted the funnel, not a second product to operate.