Google Search Console, GA4, Looker Studio and Tag Manager are free and can do a lot. But set up halfway, they give you half-true numbers. We go through how to set them up, what each tool is for and how it all ties in with cookie consent and Consent Mode.
Almost every company has some analytics on its website. Hardly any of them trust it. Numbers from ad platforms don't match the e-shop, half the orders are missing in GA4 and nobody remembers who set it up and when. The result is the same as having no analytics at all: decisions are made by gut feeling.
This article is a guide to the four tools that form the foundation of measurement on most websites. All of them are from Google and all are free. We're not writing it as a how-to for every button, as those keep changing. We're writing it so you know what each tool is for, what has to be set up and what to watch out for.
Google Search Console: setup and using the data
Google Search Console shows how Google sees your website. How many times it appeared in search results, for which queries, how many people clicked and whether Google has any technical problems with the site. It is the only place where you get this information straight from Google.
How to set up Search Console
- Add the website as a domain property. It covers the whole domain including all subdomains and both the www and non-www versions. It is verified with a DNS record, which your domain administrator can help with.
- Submit the sitemap. The address is usually /sitemap.xml. That way Google learns about all the pages you want in search.
- Add other users. A colleague, an agency, a developer. But keep ownership with the company account.
- Link Search Console with GA4, so you can see search data alongside what people do on the website.
What to track in Search Console
- Performance. The queries you appear for, impressions, clicks, click-through rate and average position. The interesting ones are mainly queries where you appear a lot but people don't click. Often all you need there is a better page title and description.
- Page indexing. Which pages Google has put in the index and which not, and why. This is where you find forgotten noindex tags, 404 errors and duplicates.
- URL inspection. Shows how Google sees a specific page and can request that it be crawled again.
- Core Web Vitals. Speed and stability of pages from the point of view of real visitors.
- Rich results. Whether Google reads structured data correctly, for example for products or FAQs.
More about what to do with the technical errors Search Console finds is in the article on SEO audits.
Detailed guide: Google Search Console: how to set it up and what to read from it.
GA4 for beginners
Google Analytics 4 (GA4) measures what people do on the website: where they came from, what they looked at, where they left and whether they bought or sent an enquiry. Unlike the older version, it is built entirely on events. A page view is an event, and so are a click, a form submission and a purchase.
What to set up right at the start
- Data retention. By default GA4 keeps detailed data for only two months. In the admin settings, extend it to 14 months, the maximum for standard properties, otherwise in a year you’ll have nothing to compare against.
- Key events. Mark the actions that have value for you: a purchase, a submitted enquiry, a booking, a phone call. These used to be called conversions.
- E-commerce tracking. For an e-shop it isn't enough to know a purchase happened. You also need the order value and the products. Without that you can't work out which channel earns money.
- Excluding internal traffic. Your own visits and those of developers shouldn't distort the data.
- Linking with Google Ads and Search Console. So you can see the whole journey from ad or search to purchase.
What to look at in GA4
Start simple. Where people come from (source and medium) and which sources bring key events. Which pages are entry pages and where people leave. How behavior differs on mobile and on desktop. For deeper questions GA4 has the Explore section, where you can build your own views, for example a purchase funnel step by step.
A few questions GA4 should be able to answer right after setup:
- Which traffic source brings the most purchases or enquiries, and which brings only visits?
- Which pages do people most often enter the website on and where do they most often leave?
- How many people add goods to the cart and how many of them complete the purchase?
- How does the conversion rate differ on mobile and on desktop?
- Do people who have already bought once come back?
If you can't find an answer to them, or don't trust it, some of the settings above are probably missing.
And one important thing: GA4 never shows a hundred percent of reality. Some people decline cookies, some have blockers, and different tools attribute a purchase to a source differently. That's why numbers from GA4, from ad platforms and from the e-shop always differ a little. The goal is not a perfect match, but to know by how much and why they differ, and to make decisions based on trends.
Step by step in the article Google Analytics 4 for beginners.
Looker Studio: reporting without Excel
Looker Studio (formerly Google Data Studio) is a tool for dashboards and reports. It can connect to GA4, Search Console, Google Ads, Google Sheets and, through connectors, other sources too. The data updates itself, so there's no more monthly retyping of numbers into Excel.
A typical marketing report in Looker Studio shows revenue and costs by channel, their ratio against the target and development over time. But a good report is above all short. It helps to first write down which decisions you want to make from it, and only then put the charts together. More on how to choose the right numbers in the article on KPIs and business dashboards.
What to put in the first report
When you're starting out, one page is enough. For example:
- Four numbers at the top for the chosen period: revenue or number of enquiries, marketing costs, their ratio (cost-to-revenue ratio or cost per enquiry) and a comparison with last year.
- A chart of development over time, ideally by week, with last year's period for comparison.
- A table of channels: where people came from, what it cost and what it brought.
- Search: queries and pages from Search Console that bring the most people.
Only once this works and decisions are regularly made from it does it make sense to add detail.
What to watch out for:
- Connecting non-Google sources (such as Meta ads) usually requires a paid third-party connector.
- E-shop revenue and costs are often not in any Google tool. The simplest way is to keep them in a Google Sheet and connect that to the report.
- Access. A report is only visible to those you share it with, but the data sources are connected under a specific account. If that account disappears, the report stops working.
How to build a report that management will actually read: Looker Studio: reporting for management.
Tag Manager and Consent Mode
Google Tag Manager and Consent Mode belong together. Tag Manager is a single code on the website that fires GA4, Google Ads, Meta and Sklik tracking. Consent Mode tells Google tags whether the visitor accepted cookies. Without consent, basic mode sends Google no data at all, while advanced mode sends only cookieless signals.
Google Tag Manager
Google Tag Manager is a manager for tracking codes. Instead of a developer putting separate code for GA4, Google Ads, Meta, Sklik and others into the website, they insert just one Tag Manager code. Everything else is then set up in its interface: tags (what should fire), triggers (when it should fire) and variables (what data is sent, such as the order value).
The upside is that you can adjust measurement without touching the website's code. The downside is that over time the container fills up with tags nobody knows the purpose of. Regular cleanup and brief documentation of what is set up and why pays off.
Cookie consent and Consent Mode
In Czechia and across the EU, analytics and advertising cookies may only be used with the visitor's consent. That's why you need a cookie banner that really collects consent, and measurement that respects it.
Consent Mode is a way of telling Google tools whether the visitor gave consent or not. The tools behave accordingly: with consent they measure normally, without it they don't store cookies. In the default variant nothing is sent without consent. In the advanced variant anonymous signals are sent, from which Google estimates the missing data. For Google ads in the EU, Consent Mode is now a condition for conversion measurement and remarketing to work.
The cookie banner, Tag Manager and Consent Mode have to work together. The most common mistake we see: the banner is displayed, but measurement runs before the visitor clicks anything. That is a legal problem and a data problem.
We cover the whole setup in the article Google Tag Manager and Consent Mode v2.
The most common measurement mistakes
We see five measurement mistakes most often: purchases counted twice, missing order values, the payment gateway showing up as the traffic source, tracking that runs without cookie consent, and nobody knowing what is set up. Each one distorts how much a channel really earns, and automated bidding in campaigns then learns from the wrong numbers.
- Purchases are counted twice, for example when the thank-you page is refreshed.
- Order value is missing, so you know how many purchases there were, but not how much they earned.
- The payment gateway shows up as the traffic source and “steals” purchases from the ad that brought them.
- Measurement runs without cookie consent, or conversely doesn't work at all with it.
- Nobody knows what is set up. Accounts on private emails, containers full of old tags, no documentation.
Where to start: measurement step by step
Start with a seven-point check: account ownership, cookie consent, key events, a check against e-shop orders, connected tools, data retention and one regular report. Every point you can’t answer “yes” to is a place where data gets lost, and where you split the budget based on numbers you can’t trust.
- Ownership. Are all accounts (GA4, Search Console, Tag Manager, ad platforms) on company accounts, and do you have admin access to them?
- Cookie consent. Does measurement only fire after consent, and is consent passed on via Consent Mode?
- Key events. Is everything that has value for the company measured: purchases with value, enquiries, bookings, phone calls?
- Check against reality. Does the number of purchases in GA4 roughly match the number of orders in the e-shop? If not, do you know why?
- Linking. Are GA4, Search Console and the ad platforms linked so they pass data to each other?
- Data retention. Is it set to the longest possible so you can compare year over year?
- Report. Is there one place where decisions are regularly made from the numbers?
Most websites we see fail at least two points. Nothing terrible, it can be fixed. It's just good to fix it before you allocate the next budget based on those numbers. And after every bigger change to the website, such as a redesign or a change of e-shop platform, the whole list needs to be gone through again. Measurement breaks most often during rebuilds.
Data means nothing without interpretation
Well-set-up measurement is the foundation, but on its own it decides nothing. Analytics shows what people do, not why, and it can’t tell whether the channel with the most purchases is the one you earn most on. That’s why we handle measurement at the second step of the 4B framework and tie the numbers to margins and goals.