Somebody clicks a link, reads three pages, leaves, comes back two days later from a different device, reads two more pages, and then finally creates an account. In your signup table, that shows up as one row with a timestamp. Everything before it is invisible unless you went and set up anonymous tracking on purpose.
I run a few small products and content sites, and the signup event is the least interesting part of the whole story. It’s the last click in a chain that started somewhere else entirely, and if I only measure what happens after signup, I’m missing most of the decision.
What “anonymous” actually means here
“Anonymous” in analytics doesn’t mean you know nothing about the person. It means you don’t yet have an identity they’ve given you voluntarily, like an email or a username. What you usually do have is a browser, a device, and a session, and most tracking tools give that combination a random ID the moment someone lands on your site.
Tools like PostHog, Mixpanel, and GA4 all do a version of this. On first visit, the script drops an anonymous ID into a cookie or local storage. Every page view, click, and event on that visit gets tagged with that ID. It’s not a person. It’s a placeholder for “whoever is holding this browser session right now.”
The honest way to think about it: you’re not tracking a person before they sign up. You’re tracking a browser’s behavior, and hoping it stays attached to the same person long enough to matter.
What actually gets captured before signup
For a typical small SaaS or content site, the pre-signup trail usually includes a handful of things:
- The landing page they hit first, and the referrer or UTM parameters that got them there
- The pages they viewed, in order, and how long they stayed on each
- Clicks on specific buttons or links, like a pricing page CTA or a demo request
- Scroll depth, if you’ve bothered to set that up
- Repeat sessions, if the anonymous ID survives long enough to see them come back
None of this requires a name or email. It requires that the anonymous ID persist across the visit and, ideally, across return visits. That persistence is the whole game, and it’s shakier than most dashboards let on.
I use this mostly for one question: what did someone actually do before they decided to trust us with an email address? For one of my sites, the answer turned out to be “read the pricing page twice before ever hitting a signup button,” which told me the objection was cost, not confusion. I wouldn’t have known that from signup data alone.
Turning an anonymous visitor into a known user
The moment someone signs up, you get to do something called identity resolution, which is just a fancy way of saying “attach the anonymous ID to the real user ID.” Most tools have an identify call for exactly this. You fire it once at signup, and every one of that browser’s earlier anonymous events gets retroactively stitched to the new user record.
This is the payoff for setting up anonymous tracking in the first place. Without it, your funnel starts at signup. With it, your funnel can start at “first touch,” which is a very different picture. You can now say a user came from a specific ad campaign, read four blog posts over a week, then signed up on their third visit. That’s the kind of thing that actually changes what you do with your marketing budget, not just what you report on it.
The catch is that this stitching only works if the anonymous ID survived intact from first visit to signup. Which brings up the part nobody likes to talk about.
Where the trail breaks
Anonymous tracking looks clean in a demo and messy in real life. A few of the ways it falls apart:
Cookie deletion and private browsing. If someone clears cookies, switches to incognito, or their browser auto-clears storage on close, they get a brand new anonymous ID next time, and your tool has no idea it’s the same person. Their “first visit” data resets.
Cross-device behavior. Someone reads your blog post on their phone during lunch and signs up from their laptop that evening. Unless they’re logged into something that bridges the two (or you’re using a paid identity resolution service, which brings its own accuracy tradeoffs), those are two separate anonymous IDs with no link between them. The laptop signup looks like a cold, direct visit even though it wasn’t.
Ad blockers and consent tools. A meaningful slice of visitors block tracking scripts outright, or decline cookies in a consent banner. Depending on your setup, that means some fraction of your traffic never gets an anonymous ID at all. Your pre-signup data isn’t a full picture of your visitors; it’s a picture of the visitors who let you track them.
Bots and scanners. Some of what shows up as “anonymous sessions” is automated traffic, security scanners, or preview crawlers from messaging apps unfurling a link. If you don’t filter for this, your pre-signup funnel looks busier than it actually is.
None of this means the data is useless. It means it’s a sample, not a census, and you should treat it that way when you’re deciding what it’s telling you.
Why I still bother with it
Even with all those gaps, pre-signup tracking answers a question that signup data alone can’t: what convinced someone, and what almost stopped them. For a one-person operation, that’s worth more than most other analytics you could set up, because it’s the difference between guessing why signups are slow and actually seeing where people drop off.
Concretely, I look at three things:
- First-touch source. Where did the anonymous session originate, so I know which channels are actually feeding signups instead of just feeding traffic.
- Pages before signup. Which pages tend to show up in the path of people who convert, versus people who bounce after one page.
- Time to signup. Whether people convert on their first visit or need multiple sessions, which tells me whether the offer is clear enough to act on immediately or needs more convincing.
None of these numbers are precise. They’re directional. If first-touch data says a channel drives signups, that’s a reasonable thing to lean into. It’s not proof that the channel caused the signup in some clean, isolated way, because attribution models are choices, not facts, and a different model can tell a different story from the same raw events.
What this data can’t tell you
It can’t tell you why someone signed up, only what they did on the way there. It can’t guarantee that improving a page that shows up often in pre-signup paths will increase signups, because correlation in a funnel isn’t the same as causation. And it can’t give you a complete picture of your traffic, because the visitors who block tracking or clear cookies simply don’t show up, or show up fragmented into multiple “different” people.
Treat pre-signup tracking as one input into a decision, not the decision itself. It’s genuinely useful for spotting patterns worth investigating further, like a page that keeps showing up right before people leave, or a channel that brings visitors who read a lot but rarely convert. What you do with those patterns still takes judgment, not just a dashboard.
If you’re setting this up for the first time, start small: get the anonymous ID persisting reliably, fire the identify call at signup so the stitching works, and resist the urge to build elaborate attribution models before you trust the basic data. Get the plumbing right first. The interesting questions come after that.
If you want more of this kind of plain-English breakdown of the metrics and methods behind running a small business on real numbers, you can find the rest of what we’ve covered on the home page.
Get new guides and videos first — join the Telegram channel.