In our last analytics post we argued that you only need three numbers to know whether your website is working: engaged sessions, key events, and traffic acquisition. If you have been reading those three for a month or two and they have started raising new questions — not “is it working?” but “why is this channel outperforming that one?” — then you are ready to go a level deeper.

This tutorial covers three advanced moves, none of which require a developer or a line of code. Each one exists to answer a sharper business question than the standard reports can. We will build up in order: filtering what you already have, then building custom analyses from scratch, then running the single most underused report in GA4.

A quick map: Standard Reports vs. the Explore Section

GA4 has two worlds. The Reports section (left sidebar) is the pre-built world you already know — fast, reliable, and fixed. The Explore section is the build-your-own world, where you assemble analyses GA4 doesn’t offer out of the box. Most owners never open Explore. That is exactly where the depth lives, and where the rest of this tutorial spends most of its time.

One rule worth learning before anything else, because it explains a lot of confusion: the tool you use to slice your data depends on which world you are in. In standard reports you use Comparisons. In Explore you use Segments. They look similar but are not equally powerful, and knowing the difference saves hours.

Move #1: Comparisons — slice a standard report without leaving it

The fastest upgrade to your weekly routine is the Comparisons feature, which lets you split any standard report into side-by-side groups. Instead of one line for “all traffic,” you can see mobile vs. desktop, or new vs. returning visitors, or one country vs. another — all in the same familiar report.

Where to find it: open any standard report (Traffic acquisition is a good one) and click Add comparison + at the top. Build a condition such as Device category = mobile, and GA4 draws a second data series alongside your total.

Why it matters for a business: your engagement rate might look healthy overall while quietly hiding a problem. Split it by device and you may discover mobile visitors are engaging at half the rate of desktop — a clear signal that your mobile experience is costing you leads. The overall number never would have told you that.

Connect4 Tip: Comparisons are quick but deliberately limited — they cannot do sequences (“visitors who did X and then Y”). The moment your question involves an order of events, you have outgrown Comparisons and it is time to move into Explore, which is Move #2.

Move #2: Explorations — build the analysis GA4 doesn’t ship

The Explore section (left sidebar, the compass icon) is a blank canvas for custom analysis. You drag in the dimensions, metrics, and segments you care about and GA4 assembles the report live. Two exploration types earn their keep for most businesses.

The Funnel exploration shows you where people drop off in a sequence you define — for example, viewed services page → viewed contact page → submitted form. If 200 people reach step two and only 12 complete step three, you have found exactly where your leads are leaking. As of 2026 GA4 supports both open and closed funnels, so you can allow people to enter mid-sequence or require they start at the top.

The Free-form exploration is a flexible table where you choose the rows and columns — think of it as a pivot table for your website. Landing pages as rows, key events as the metric, broken out by channel: a view no standard report gives you in one place.

Segments are the payoff here. Unlike Comparisons, Explore lets you build Segments that can include sequences and time constraints — “users who visited a blog post, then returned within 7 days and submitted a form.” That precision is only available in Explore, which is the main reason to work here at all.

Connect4 Tip: Explorations sample your data once a date range covers roughly 10 million events — standard reports never sample. For most small businesses you are nowhere near that ceiling, so trust your Explore numbers. But if you ever analyze a very long date range and see a “this report is based on a sample” notice, shorten the range and the numbers sharpen back up.

Move #3: The Retention Report — which channels buy customers, not just clicks

This is the report almost nobody opens, and it quietly reframes the “which channel is winning?” question from our first post. Traffic acquisition tells you which channel brought the most engaged sessions this week. Retention tells you which channel brought people who came back.

Where to find it: Reports → Retention. It shows how returning users behave over the 42 days after they first arrive. The move that pays off is comparing retention across acquisition channels — and the result is often counter-intuitive.

A common pattern: paid social ranks high on first visit and low on 30-day return, while organic search and email do the opposite — fewer people up front, but far more who come back. A last-click report would tell you to pour budget into paid social. Retention tells you that organic and email are the channels actually building a durable audience. That gap is often exactly where a budget reallocation should happen.

Connect4 Tip: Read retention by acquisition channel once a month, not weekly — it needs time to accumulate meaning. It is the clearest single view of the difference between traffic that buys a click and traffic that buys a customer.

One “advanced” feature to be skeptical of (for now)

You will see other guides push GA4’s predictive audiences — machine-learning segments like “likely to purchase” or “churn risk.” They are genuinely powerful, but they carry a hard requirement most small businesses can’t meet: GA4 needs at least 1,000 returning users who triggered the relevant event, and 1,000 who did not, within the prior 28 days, before it will generate the prediction.

Below that volume, the feature simply reports “insufficient data” and never populates. That is the single most common reason small-merchant predictive audiences fail. So if your site does not yet see that kind of returning traffic, skip predictive audiences for now, focus on the three moves above, and revisit prediction once your volume grows. We would rather tell you that plainly than send you chasing a feature that won’t work yet.

What You Can Do Right Now (No Developer Needed)

• Open Traffic acquisition, click Add comparison, and split it by Device category. Note whether mobile engagement lags desktop — if it does, that is your next website priority.
• In Explore, build one Funnel exploration for your main conversion path (e.g., services → contact → form submit) and find the biggest drop-off step.
• Build one Free-form exploration with landing pages as rows and key events as the metric, to see which pages actually produce leads.
• Open Reports → Retention and compare how your top two acquisition channels perform over the 42 days after first visit.
• Before trusting any long-range Explore report, check for a sampling notice. If you see one, shorten the date range.

How Connect4 Can Help

Explore is powerful precisely because it is open-ended — which also means it is easy to build a report that looks authoritative but measures the wrong thing. Connect4 folds this deeper analytics work into an integrated monthly care plan alongside security, performance, content, and visibility, so the numbers you act on are the right ones.

• Build and save the Explorations that matter for your business — a lead funnel, a landing-page performance table — so they are one click away, not rebuilt each time.
• Define the Segments and key events that make funnels accurate, including the sequence-based segments that Comparisons can’t handle.
• Set up a simple Looker Studio dashboard that surfaces your Explore insights and retention trends on one screen, no navigation required.
• Interpret the retention-by-channel picture and translate it into a concrete budget-reallocation recommendation.
• Roll analytics into a monthly care plan so measurement is maintained alongside security, performance, and content — not rebuilt from scratch each quarter.

Frequently Asked Questions

What is the difference between Comparisons and Segments in GA4?

Comparisons filter standard reports and are quick but limited — they cannot handle sequences of events. Segments live in the Explore section and are far more flexible, supporting sequences and time constraints. Rule of thumb: use Comparisons for a fast split in a standard report, and switch to Explore with Segments the moment your question involves the order in which things happened.

What is a Funnel exploration used for?

A funnel exploration shows where people drop off in a step-by-step path you define, such as viewing a services page, then a contact page, then submitting a form. It pinpoints the exact step where potential leads are leaking, so you know which page to fix. GA4 supports both open funnels (people can enter mid-sequence) and closed funnels (they must start at step one).

Why does my GA4 Exploration say the data is sampled?

Explorations apply sampling once a date range covers roughly 10 million events, while standard reports never sample. Most small businesses are well below that threshold, but a very long date range can trip it. If you see a sampling notice, shorten the date range and the numbers become exact again.

Can I use GA4’s predictive audiences for my small business?

Only if you have the volume. GA4 requires at least 1,000 returning users who triggered the relevant event and 1,000 who did not, both within the prior 28 days, before it will generate a prediction. Below that, the feature shows “insufficient data.” Many small businesses aren’t there yet, so it is usually better to master Explorations and retention first and revisit prediction as traffic grows.

What is the retention report and why should I care?

The retention report shows how returning users behave over the 42 days after they first arrive. Comparing it across acquisition channels reveals which channels bring people who come back — not just people who click once. It frequently ranks your channels differently than a last-click report does, which is exactly where smarter budget decisions come from.

The Difference Between Watching Numbers and Asking Questions

The three numbers from our first post tell you whether your website is working. The three moves in this one let you ask why — and why is where decisions actually get made. Comparisons split a report open, Explorations build the view GA4 never shipped, and the retention report tells you which traffic is worth paying for. None of it needs code, and all of it compounds. And when an Exploration raises a question you would rather hand off than chase, that is exactly what a monthly care plan is for.