Blog / Free Advice and Other Musings

A practical GA4 guide for marketers who want to know what’s working, what needs attention, and what to do next.

Most analytics dashboards are pretty good at telling you what happened: Traffic went up. Organic Search brought in this many visitors. A product page received 2,000 views. Fifty people added something to their cart. The harder observation is, so what:

What should marketing do about it?

That is where analytics often falls short. For many marketers who do not spend a lot of time in Google Analytics, the reports are obtuse tables of channels, events and conversion rates. It’s hard to divine a strategy from it. For those familiar with data, analytics is like looking at a chessboard. You see the position, the pressure points, and the opportunities. The value is not the data itself, it’s what drives the move you make next.

Novice chess players learn to master the game by starting with the basics, focusing on core opening moves, understanding the middle game, and reviewing every game for mistakes. Maybe we can take a similar approach with analytics by starting with the basic business questions, focusing on where to start, understanding the “middle”, and creating some strategies for deciding what action each signal suggests.

Here are the six basic questions I begin with when reviewing a client’s analytics:

1. Can people find us?
2. Are we attracting the right visitors?
3. Are visitors finding what they need?
4. Which website activities create value?
5. Where should we invest more?
6. What should we fix?

One important caveat before we start: Google Analytics tells you what people do after they reach your website and consent to tracking. It cannot tell you everything. To get more of the picture, use Google Search Console to understand visibility in Google Search, use GA4 to understand what happens when visitors arrive, and use the website itself, user feedback, and testing to understand why people behave the way they do.

Let’s look at how we best answer those 6 questions, keeping in mind the best starting point, the ways to interpret the data, and the strategic action to take depending on the signals in the data.

Part A: Discovery (the opening moves)

1. Can people find us? Search visibility and website traffic

A common analytics question is:

How much traffic did we get?

That is usually too vague. A better question is:

Are more or fewer people finding us, and where are they finding us?

There are really two questions hiding in there: what’s your visibility in search and are people clicking through to the site.

To answer, are people seeing you in Google Search, start with Google Search Console and look at:

• Search impressions
• Search clicks
• Search queries
• Pages appearing in search results
• Click-through rate

Signal/Action: If impressions are declining, you may have a visibility problem. If impressions are rising but clicks are not, the issue may be how your pages appear in search results, or whether the page actually matches what people are looking for.

To answer, are people reaching the website, go to GA4 and look at: Reports → Acquisition → Traffic acquisition

Compare channels such as:

• Organic Search
• Direct
• Referral
• Email
• Paid Search
• Organic Social
• Paid Social

First, look at traffic volume. Then look at what those visitors do.

A traffic source that sends a lot of visitors but almost no purchases tells you something very different from a smaller source that consistently produces cart activity and revenue. For example, many of the sites I review have more traffic from Organic Social than from Email, but Email often produces more add to carts and more purchases.

Turn the numbers into a diagnosis

Signal	What it might mean	What to do next
Search impressions are down	Search visibility may be declining	Check affected pages, indexing and SEO
Impressions are high but clicks are weak	Search listings may not be persuasive or relevant	Review page titles, descriptions and search intent
Traffic is down but conversion quality is steady	Acquisition problem	Work on reach
Traffic is up but valuable actions (key events) are weak	Traffic-quality or landing-page problem	Examine source, targeting and landing pages
Traffic and valuable actions (key events) are both rising	Something is working	Identify the source and support it

The point is not simply to report that traffic was up or down. The more strategic question is, “Where did the change happen, and does it matter?”

Watch the video walk through on using the Traffic Acquisition report to answer the above questions.

Q2. Are we attracting the right visitors? Measure traffic quality in GA4

More traffic is not automatically better traffic. Suppose a paid social campaign drives a large increase in visits to your site. That sounds good. But now compare those visitors with people arriving through Organic Search. (I like using Organic Search as the comparison because it’s usually the highest quality traffic coming to any given site.)

Which group is more likely to:

• View products?
• Add products to the cart?
• Begin checkout?
• Purchase?
• Generate revenue?

The key starting point is:

Which sources bring visitors who actually do something valuable?

Step 1: Look at sessions. Sessions tell you about reach. Which channels are bringing people to the site?

Step 2: Look at what those visitors do. Depending on the business, useful measures might include:

• Product views
• Add-to-cart activity
• Ecommerce purchases
• Revenue
• Key events
• Engaged sessions

Step 3: Compare rates, not just totals. Large traffic sources tend to generate more conversions simply because they have more visitors.

So also ask:

What percentage of visitors from this source did something useful?

That gives you a better picture of traffic quality. But remember to define what “useful” means. Do not judge every source against the same outcome. An awareness campaign and a sales campaign may have completely different jobs.

For example:

Awareness campaign

• Reach
• Engaged visits
• Product or content views

Sales campaign

• Add to cart
• Checkout
• Purchase
• Revenue

Subscriber campaign

• Newsletter signups

Before judging a traffic source or campaign, ask:

What was this traffic supposed to do?

Otherwise you risk calling a campaign unsuccessful simply because you are measuring it against the wrong goal.

Part B: Engagement (the messy middle)

Q3. Are visitors finding what they need? Use GA4 to identify content and navigation problems

This is a question analytics cannot answer perfectly. GA4 cannot tell you: “Yes, this person found exactly what they wanted and left satisfied.” But it can provide useful signals.

I like to look at:

• Landing pages
• Pages and screens
• Product views
• Site search
• Important interactions
• What visitors do next (path explorations)

The goal is to understand whether people are reaching relevant content and whether that content helps them continue.

Start with your site search. If someone searches your website, they are explicitly telling you what they came looking for.

Use the site search report (usually found under Engagement) and answer:

• What are the most common searches?
• Which searches produce little or no useful content?
• Do visitors immediately search again?
• Are people repeatedly searching for something that should be easier to find through navigation?

Repeated searches do not automatically mean something is wrong. But they can point toward content that is:

• Difficult to find
• Poorly labelled
• Missing
• Buried too deeply in the site

For an ecommerce site, visitors may repeatedly search for a product category, size, brand, or feature. For a publisher, they might search for an author, exam copy information, permissions, or a particular subject. Those searches are useful clues for improving navigation and content.

TIP: here’s how to set up site search in GA4

To take that information further: Look at the visitor’s next step. A useful page should generally help someone move forward.

For an ecommerce site: Product view → Add to cart → Purchase

For another type of organization it might be:
Article → Newsletter signup
Event page → Registration
Book page → Table of contents → Exam-copy request

Signal/Action: If a page gets substantial traffic but very little next-step activity, that does not automatically mean the page is bad. It tells you: This page is worth investigating.

Analytics identifies where to look. It rarely tells you the complete reason why something is happening.

Watch the video walkthrough on how to go beyond the traffic report

Q4. Which website activities actually create value?

For ecommerce, the economic outcomes are fairly obvious:

• Purchases
• Items purchased
• Revenue
• Average order value

You can explore product-level performance in: Reports → Monetization → Ecommerce purchases

But many organizations have more than one valuable website outcome.

A publisher might care about:

• Purchases
• External retailer clicks (ShopLocal)
• Newsletter signups
• Exam-copy requests
• Media-copy requests
• Catalogue downloads

An arts organization might care about:

• Ticket purchases
• Memberships
• Donations
• Registrations
• Newsletter signups

A professional organization might care about:

• Membership inquiries
• Downloads
• Webinar registrations
• Contact requests

Not every action is of value. Instead consider:

Which business outcome are we trying to increase?

Then work backwards.

Example: If you want to increase purchases then ask:

1. How many purchases occurred?
2. Which products were purchased?
3. Which traffic sources produced them?
4. Which campaigns produced them?
5. Which landing pages were involved?
6. Which products were viewed or added to cart frequently but rarely purchased?

That last question can expose opportunities that a simple revenue report misses.

The same process works for any outcome.

Example: If you want more newsletter subscribers:

1. How many subscriptions occurred?
2. Which pages produced them?
3. Which channels brought those visitors?
4. Which campaigns generated subscriptions?
5. Which high-traffic pages produce very few signups?

Signal/Action: Start with the result you want to increase. Then use analytics to understand what pages or events tend to lead to that desired interaction.

Watch the demo on using the Path Exploration to understand what pages or events lead to a final outcome

Part C: What should marketing do next? (strategic moves based on the signals)

Q5. How to find marketing opportunities in GA4

This is where analytics starts becoming genuinely useful. One of the simplest ways to review campaigns, landing pages, or products is to compare two things: How much attention did it get? and How often did that attention turn into useful action?

You can think about it as a simple two-by-two matrix.

	Low action rate	High action rate
High traffic	Fix	Support
Low traffic	Test or deprioritize	Grow

This same framework can be used for:

• Campaigns
• Traffic sources
• Landing pages
• Products
• Books
• Content
• Email promotions

High traffic + high action: support it
This is a signal that things are working. Ask whether you can increase:

• Budget
• Frequency
• Email exposure
• Internal links
• Organic promotion

The important thing is to understand why it works so you can repeat it.

Low traffic + high action: grow it
This is often the most interesting category. The people who reach the page, product or campaign respond well. The problem may simply be that not enough people see it.

Ask: Can we get this in front of more of the right people?

High traffic + low action: investigate
This deserves attention because a lot of people are seeing it. Possible causes include:

• Poor targeting
• A mismatch between the campaign message and the landing page
• Weak calls to action
• Page usability
• The wrong offer
• The wrong campaign goal

Do not immediately assume the page itself is the problem. Trace the full journey.

Low traffic + low action: test or deprioritize
Usually this is not where I would spend hours optimizing. Marketing teams have limited time. Analytics should help you decide where your attention is most likely to matter.

Use the same method to find products worth promoting

The most-viewed product is not necessarily your biggest marketing opportunity. For ecommerce, you can compare product metrics such as:

• Items viewed
• Items added to cart
• Items purchased
• Item revenue

Instead of asking:

Which products get the most views?

ask:

Which products perform unusually well once people find them?

Think about products in the same matrix:

	Low purchase intent	High purchase intent
High visibility	Investigate	Keep supporting
Low visibility	Lower priority	Promotion opportunity

Imagine Product A gets a huge number of views but relatively few people add it to their cart. Product B gets far fewer views, but a much larger proportion of those visitors add it to their cart or purchase it.

Product B may deserve:

• An email feature
• More internal links
• Social promotion
• Paid advertising
• Better placement on a category page

Signal/Action: Instead of simply ranking products by pageviews, the observation should lead to an action.

This product performs well once people find it. Can we help more people find it?

Tip: Be careful with tiny samples. One purchase from three product views does not make something your breakout product. Set a reasonable minimum amount of activity before treating an unusually high conversion rate as meaningful.

Think of outliers as signals to investigate, not proof.

Q6. How to use GA4 to diagnose a website or marketing problem

Sometimes the analytics problem is not: Where is the opportunity? It is: What went wrong?

Use the same diagnostic sequence every time.

Step 1: What changed? Was it:

• Traffic?
• Product views?
• Cart activity?
• Purchases?
• Revenue?
• Signups?
• Registrations?

Start with the actual change.

Step 2: Where did it change? Break the number down by:

• Channel
• Campaign
• Landing page
• Product
• Device
• Geography, if relevant

Usually the overall number is not the most useful part. The segment where the change occurred is.

Step 3: Was the change intentional? Did you:

• Launch a campaign?
• Send an email?
• Receive media coverage?
• Promote a particular product?
• Change the website?
• Publish new content?

A spike is not mysterious if you caused it.

Step 4: Did the rest of the visitor journey change too?

Compare: Traffic up → Product views up → Purchases up with Traffic up → Product views up → Purchases DOWN
Those patterns suggest very different problems.

Likewise: Traffic DOWN → Purchase rate steady may indicate that the site itself is working fine.
The problem may simply be that fewer people are arriving.

Step 5: Could the tracking be wrong? Always ask this before making an expensive business decision. Broken tracking has caused many imaginary marketing crises. Check:

• Did analytics tags change?
• Did consent settings change?
• Did a checkout or form change?
• Did the event stop firing?
• Did the website launch something new?

Verify the measurement before diagnosing the marketing.

Step 6: Look at the actual page

Analytics tells you where to look. It cannot usually tell you whether the underlying problem is:

• Confusing copy
• A buried button
• A broken form
• Weak imagery
• Poor navigation
• A mismatch between visitor expectations and the page

Open the page. Use your judgement.

Step 7: Make one change and measure again. Analytics should be a feedback loop:
Observe → Investigate → Change → Measure

Not: Report → File PDF → Repeat next month

---

A 20-minute bi-weekly Google Analytics routine for marketers

You do not need to spend half a day in Google Analytics every month. For most small marketing teams, a short structured review is more useful than aimless exploration.

1. Visibility and traffic — 5 minutes

Ask: Are people finding us?

Look at:

• Search visibility
• Overall website traffic
• Organic traffic
• Major traffic channels

Look for significant changes. Do not investigate every tiny fluctuation.

2. Outcomes — 5 minutes

Ask: What useful things happened?

For ecommerce that might include:

• Purchases
• Revenue
• Cart activity

For your own organization, include the outcomes that actually matter. That could include:

• Newsletter signups
• Registrations
• Donations
• Downloads
• Inquiries
• Exam-copy requests

3. Opportunities — 5 minutes

Look for:

• Products with strong purchase intent but limited exposure
• Campaigns with unusually strong response
• Landing pages performing particularly well
• Organic-search pages attracting useful traffic
• Traffic sources producing high-value visitors

Choose 1 to 3 opportunities worth acting on.

4. Problems — 5 minutes

Look for:

• Sudden traffic loss
• High traffic with weak action
• Campaigns producing visits but little else
• Heavily viewed products with unusually weak purchase intent
• Internal searches suggesting people cannot find something
• Important pages receiving little use

Again, choose the few things worth investigating. The goal is not to leave your monthly analytics review with 37 observations. The goal is to leave with a short list of things to do.

Turn analytics observations into marketing actions

A useful review of your analytics should eventually lead to something like this:

Observation	What it might mean	Action	What to watch
Organic traffic converts strongly	Search is bringing qualified visitors	Support SEO and strong landing pages	Organic traffic + purchase rate
Product has modest traffic but strong cart activity	It may be under-promoted	Feature it in email or social	Product views + cart activity
Campaign brings lots of visitors but few useful actions	Targeting, message or landing page may be weak	Review the full campaign journey	Key-event or purchase rate
Important page gets little use	Visitors may not find it or need it	Review placement and usefulness	Page views + next-step behaviour
Site search repeatedly includes the same topic	Content may be hard to find	Improve navigation or create clearer content	Search frequency + resulting page visits

This is the step I wish more marketers would take. The dashboard or set of analytics reports you review should not be a wall of numbers. It should look like clues telling you want to do next, how to plan for the next week, 28 days, quarter, etc.

In summary, analytics should answer “What should we do next?”

Google Analytics can provide an extraordinary amount of data. You do not need most of it.

1. Start with the business question
2. Find the relevant signal
3. Decide whether it represents a problem or an opportunity
4. Take an action
5. Measure whether the action helped

That is the useful analytics habit: Question → Signal → Interpretation → Action → Measurement

The dashboard is evidence. The endgame is better marketing decisions.