---
title: "Analytics before launch: the minimum that makes a site readable"
canonical: "https://en.prostor-agency.com/blog/analytics-before-launch"
summary: "What to install before a site goes live, which events deserve to be goals, and why the first month of traffic is the one you cannot get back."
publishedAt: "2026-10-15"
---

## Why this happens before launch, not after

Analytics goes last because it changes nothing on screen. It is also the omission
teams regret most. The first weeks after launch are the most informative period you
will ever have — that is when people arrive without being primed by a campaign. That
data cannot be reconstructed later. No counter in week one means week one is simply
gone.

There is a practical argument too. Before launch the developer still has the code
open and can mark up the actions that matter properly. Afterwards it is done in a
hurry, on top of a live site, and the result is the kind of tracking nobody trusts
six months later.

## The minimum worth installing

**A counter with your own traffic filtered out.** On a young site your team, your
contractor and automated speed checks account for a real share of sessions. Without
a filter the dashboard reflects your own activity rather than your customers'.

**Goals on every way of making contact.** A submitted form, a messenger tap, a phone
tap, a booked call — separate goals, not one. These are different conversations with
different quality, and merging them removes your ability to tell which channel
produces meetings and which produces clicks.

**Page context on every enquiry.** Send the originating page with the event. It is
the only way to learn later which material actually brings people to a conversation
— the case study, the service page or the pricing section.

**Load speed.** Not for the report, but as an early warning. If a release doubles
the time to first paint on mobile, you want to hear it from data rather than from a
client. The link between speed and lost enquiries is covered in
[website speed and buyer drop-off](/blog/website-speed-buyer-drop-off).

## What the minimum should not include

Heatmaps, session recordings and multi-step funnels are useful tools, but not at
launch. While traffic is small they produce attractive pictures and no significance:
three recordings are three people, not a finding about your audience.

The same goes for goals like "viewed the services page" or "scrolled to the bottom".
They are events, not results. Once a dozen of those exist, the conversion number in
your report stops meaning anything, because it is measured against actions nobody
pays for.

## Connecting events to money

A counter shows enquiries, not revenue. The bridge is built by hand, and without it
analytics stays a story about clicks.

First, pass the source with the enquiry — page, form type, campaign. Then every deal
in the CRM has an origin, and a quarter later you can see which pages produce
conversations.

Second, reconcile. Once a month, compare goal counts in the analytics tool with the
number of enquiries in the CRM. A gap of several times over is almost always a
technical loss: the form submits but the event does not fire, or the notification
lands in spam. Nothing reports these on its own.

## A launch-day checklist

- The counter responds on every main page, not only the homepage.
- A test enquiry arrives and registers as a goal.
- Messenger and phone taps are counted separately.
- Your own traffic is filtered — verify it on yourself.
- Every connected service is named in the privacy policy.

The last one is not paperwork. If a tool is running and the policy does not mention
it, the site is doing something it does not admit to.

## What to look at a month in

Wait for a month of clean data before drawing conclusions. Three things are enough:
how many visits reach an enquiry, which pages those enquiries come from, and where
people leave most often. That is a basis for deciding what to fix next, instead of
guessing from impressions.