Your Analytics Setup Checklist

Stage 1 — Install and verify the foundations

Nothing else in this checklist matters until the basic tracking is confirmed working, not just installed. Work through these in order, and do not move to the next stage until each item is genuinely verified — not assumed:

• Create a GA4 property and web data stream, with the correct time zone, currency, and business details filled in (Chapter 2).
• Install the Google tag site-wide — directly, or preferably through Google Tag Manager — and confirm it fires on every page, not just the homepage.
• Open the Realtime report and browse your own site in another tab to verify the tag is actually working before doing anything else.
• Check your page source or a tag-checking extension to confirm exactly one GA4 tag is firing per page — not two, which quietly doubles every number downstream (Chapter 8).
• Set up an internal traffic filter so your own visits, and your developer's or agency's, do not blend into customer data (Chapter 8).
• Verify your Google Search Console property, using domain or URL-prefix verification depending on your setup (Chapter 4).
• Submit your XML sitemap inside Search Console, and check the Coverage report for any existing indexing errors.
• Link Search Console and GA4 together inside GA4's Admin settings so Search Console data becomes visible alongside GA4 reports.

Stage 2 — Define what actually counts as success

• Write down, in plain language, the two to five actions that represent real business value on your site — a form submission, a call click, a booking, a purchase (Chapter 5).
• Separate that list into primary conversions (the real business outcome) and micro-conversions (meaningful steps that are not the finish line on their own), so neither gets mistaken for the other later.
• Set up each primary conversion as a GA4 event (using an existing automatically-collected event where possible, a custom event via Google Tag Manager where not), and mark it as a key event.
• Test every key event yourself and confirm it appears correctly in Realtime and, the following day, in standard reports.
• If you run Google Ads, link your Ads account to GA4 and import only the genuine key events as conversions — not every event in the account.
• If any pages or forms live on a different domain or subdomain (a hosted checkout, a booking widget), configure cross-domain measurement so those journeys are not split into two disconnected visits.

Stage 3 — Build visibility and make it a habit

• Build a simple Looker Studio dashboard with five to eight metrics tied to real decisions — not a wall of every number available (Chapter 6).
• Separate it into an exec-summary view (a handful of scorecards and one trend chart) and a detailed view (channel and page-level breakdowns).
• Tag every outbound campaign link — social posts, email newsletters, paid ads — with UTM parameters before it goes live, so it shows up correctly in reports rather than as generic "referral" or "direct" traffic (Chapter 7).
• Agree on a simple, consistent UTM naming convention in writing before tagging anything, so campaigns do not fragment into inconsistent rows across reports.
• Put a recurring appointment on the calendar — monthly is usually the right cadence for a small business — to actually open the dashboard, not just build it once and forget it.
• At each monthly review: check for new Search Console indexing errors, confirm conversions are still tracking correctly, note which channels are producing results, and decide on one or two changes for the coming month based on what the data actually shows.
• On a quarterly basis, revisit the key event list and the dashboard metrics themselves — business priorities shift, and what mattered most a year ago is not guaranteed to still be the right thing to track today.

A note on sequencing

If you are setting all of this up for the first time, resist the urge to jump straight to Stage 3. A beautiful dashboard built on top of an unverified GA4 install, with no agreed definition of what counts as a conversion, is worse than no dashboard at all — it will confidently show wrong numbers, and wrong numbers dressed up nicely are more dangerous than no numbers, because they invite trust they have not earned. Each stage above depends on the one before it being genuinely correct, not just technically complete.

What "good" looks like once this is done

You do not need a perfect analytics stack to get real value from it — you need a correctly firing GA4 property, a verified Search Console connection, a short list of key events that map to genuine business outcomes, a dashboard you actually open, and a habit of reviewing it on a set schedule. That combination, done consistently, is what separates a business making decisions on evidence from one making decisions on impressions — which is exactly the gap this course set out to close in Chapter 1.

From here, the work is mostly maintenance and refinement: revisiting your key events as the business changes, tightening the dashboard as you learn which numbers you actually act on, and staying alert to the setup mistakes covered in Chapter 8 whenever the site itself changes. None of it needs to be perfect on day one — it needs to keep being checked, corrected, and used, which is the habit that actually compounds into better decisions over time.

If you have worked through all three stages above, you have already cleared the standard set out in Chapter 1: a real number for last month's conversions, a rough sense of which channel produced each one, a set schedule for checking, and at least one decision you can trace directly back to what the data showed. That is a genuinely solid foundation — the kind most small businesses never quite get around to building, even though none of the individual steps were especially difficult on their own.

As the business grows, expect this checklist to need revisiting rather than treating it as a one-time project. A new product line may need its own key event. A new paid channel needs its own UTM convention agreed before the first campaign goes live. A team member leaving may mean the person who used to check the dashboard monthly is no longer around to do it, and the habit needs to be handed to someone else deliberately rather than quietly lapsing. None of this is a sign the original setup was wrong — it is simply what "maintaining" analytics looks like once the initial setup is done.