Most GA4 implementations are configured to count pageviews. Yours should be configured to answer the questions your leadership team asks every Monday morning.
You installed GA4. The tracking code fires. The real-time view shows visitors. You have data. Except you don’t have anything useful.
The default GA4 installation gives you pageviews, session counts, user demographics, and a collection of automatically tracked events — scrolls, outbound clicks, file downloads. These are activity metrics. They tell you what happened on your website. They don’t tell you anything about your business.
When your CEO asks “Which channels are driving demo requests?” or your VP of Sales asks “Are prospects from paid search more likely to convert than organic?” — the default GA4 setup can’t answer either question. Not because GA4 can’t do it, but because nobody configured it to.
This is a configuration problem, not a platform problem. GA4 is capable of answering serious business questions. But it requires intentional setup — event architecture, custom dimensions, conversion definitions, and a data layer that connects website behavior to revenue outcomes. This is part of a broader marketing analytics and attribution framework that mid-market product companies need to get right.
The first mistake is opening GA4 and asking “What should we track?” That’s backwards. Start with the questions your business needs answered, then work backward to the data required.
Before touching GA4, sit down with your leadership team and ask: “What are the five questions you wish you could answer about our website and marketing right now?” Typical answers from mid-market product companies ($3M–$50M):
These questions define your measurement plan. Every event, dimension, and conversion you configure in GA4 should serve at least one of these questions. If a tracking element doesn’t connect to a business question, it’s noise.
Create a simple spreadsheet: Column A is the business question. Column B is the metric that answers it. Column C is the GA4 event or report that produces that metric. If you can’t fill in Column C, that’s your implementation gap. This document becomes the spec for your GA4 setup and the reference for anyone who touches the configuration later.
GA4 is event-based, not pageview-based. This is its biggest strength and the thing most implementations waste. Events in GA4 are flexible — you define them, name them, attach parameters to them. The problem is that flexibility without a plan produces a mess of inconsistently named events that nobody can analyze.
Structure your events into three tiers based on business significance:
Use a consistent naming pattern: category_action_detail. For example: form_submit_demo, content_view_case_study, cta_click_pricing. Inconsistent naming — mixing camelCase with snake_case, or using vague names like “event1” — makes analysis impossible six months later when nobody remembers what each event means.
Your event architecture should mirror your funnel. Tier 1 events are bottom-of-funnel conversion actions. Tier 2 events are mid-funnel engagement signals. Tier 3 is everything else. If you can’t place an event in one of these tiers, you probably don’t need it.
Default GA4 dimensions — source, medium, campaign, page title — answer generic questions. Custom dimensions answer your questions. They attach business context to every event, turning raw data into something your team can act on.
GA4 allows 25 custom user-scoped dimensions and 50 event-scoped dimensions. That sounds like a lot until you start adding everything you can think of. Be selective. Each custom dimension should connect to a business question from your measurement plan. If it doesn’t, it’s consuming a slot you might need later.
GA4 lets you mark any event as a conversion. Most implementations mark too many events or too few. Mark too many and every report is diluted with low-value actions. Mark too few and you’re flying blind on what drives revenue.
A conversion should represent a meaningful business outcome — an action that moves a visitor closer to revenue. For most mid-market product companies, conversions fall into three categories:
That’s typically 4–6 conversion events. Not 15. Not 2. Enough to see the full picture without drowning in noise.
Every conversion should have a monetary value, even if it’s estimated. A demo request might be worth $500 based on your average close rate and deal size. A newsletter signup might be worth $5. These values let GA4 calculate return on ad spend and channel value — otherwise you’re comparing raw conversion counts across channels that have wildly different conversion quality.
GA4 data in isolation is website data. The real value comes when you connect it to business outcomes — revenue, pipeline, customer lifetime value. This is where the analytics and attribution framework comes together.
The most impactful connection for mid-market companies is GA4 to CRM. When a lead submits a demo form, GA4 knows their entire website journey: which channel brought them, what pages they viewed, what content they engaged with. Your CRM knows the outcome: did they become a customer? What was the deal value?
Connect these two and you can answer the question every marketing team struggles with: “Which marketing activities actually produce revenue?” Not just leads. Revenue.
Don’t try to build the full integration stack on day one. Start with passing the GA4 client ID into your primary conversion form. That single connection lets you match website behavior to CRM outcomes for your most important conversion action. Expand from there once the first connection is validated and producing insights.
After configuring GA4 for dozens of mid-market product companies, the same mistakes appear repeatedly. Avoiding these saves months of bad data and misguided decisions.
More data is not better data. Every unnecessary event adds noise, complicates reports, and increases the chance of hitting GA4’s 500-event-name limit. Track what connects to a business question. Delete everything else.
Your team visits your website constantly. If you don’t filter internal IP addresses and internal users, your data is contaminated. On a smaller mid-market site, internal traffic can represent 10–20% of total sessions. That’s enough to distort every metric.
If your marketing site is on www.example.com and your app or checkout is on app.example.com, default GA4 treats these as separate properties. A user who visits your marketing site and then signs up on your app looks like two different users. Cross-domain tracking configuration is required to see the full journey.
Firing events directly from Google Tag Manager triggers based on CSS selectors and click classes is fragile. When your developer redesigns a button, the tracking breaks. A data layer — a JavaScript object that passes structured data to GTM — decouples tracking from design. It takes more upfront work but eliminates the maintenance nightmare of constantly fixing broken triggers.
GA4 configuration isn’t a one-time project. Your business questions change, your website evolves, your product adds features. Schedule a quarterly review: Are the events still firing correctly? Do the conversions still reflect the actions that matter? Are the custom dimensions still relevant? A 2-hour quarterly audit prevents the slow drift from useful data to unreliable data.
GA4 is a tool, not a strategy. Start with 5 business questions. Build a measurement plan that answers them. Configure events, dimensions, and conversions to serve that plan. Connect GA4 to your CRM so website data becomes revenue data. Then maintain it quarterly. That’s the difference between analytics that collects dust and analytics that drives decisions.
30-minute discovery call. We’ll map your stack and show you where the margin is hiding.
Schedule Call