Blog · Analytics · GA4 · Tracking · Campaign Strategy

Why GA4 setup is critical for modern campaigns

Running ads without a proper GA4 setup is like driving at night with the headlights off and the GPS set to "vibes." You'll definitely end up somewhere. Here's why the boring part of marketing is now the part that decides whether your budget gets smarter or just gets spent.

Author
Mahdi Hasan
Published
07 Oct 2026
Read time
8 min (faster than an Exploration loads)
Category
Analytics

Four numbers every GA4 owner should know

2 modefault event-data retention in a new GA4 property. Yes, two.
14 mothe maximum on a standard property, if you remember to change it
>70%of consent-lost click-to-conversion journeys Google says modeling can recover
0Universal Analytics reports you can still open. It's gone. Let it go.

Sources: Google Analytics Help · Google Ads blog (full list at the end)

The problem

"Installed" and "set up" are not the same thing

Almost every business I audit has GA4. Very few have GA4 set up. The tag is on the site, the dashboard shows a number going up and down, and everyone agrees it's "working." Then somebody asks a simple question, like "which campaign brought in our best customers last quarter?", and the room starts staring at the ceiling.

Pasting the tracking code is not a setup. It's a hello. A real setup decides which actions count as success, how much each one is worth, where it came from, how long you keep the data, and whether your ad platforms ever hear about any of it.

And here's why it matters more now than ever: modern campaigns run on automation. Smart Bidding, Performance Max and value-based bidding don't read your mind. They read your conversion data. Feed them junk and they'll optimise for junk with impressive efficiency.

Plain truthThe algorithm is a very fast, very obedient intern. It will do exactly what your tracking tells it to. That's the good news and the bad news.

The framework

How a good GA4 setup turns into smarter ad spend

Think of GA4 as the nervous system of your marketing. Ads are the muscles. Without the nerves, the muscles just twitch expensively.

Step 01 · Collect

Capture what people actually do

Deploy through Google Tag Manager, track meaningful events (sign-ups, add-to-cart, purchases, calls) with clean, consistent names and parameters.

Tools: GTM, enhanced measurement, data layer
Step 02 · Define

Decide what "success" means

Mark only the actions that matter as key events, and attach real values: order revenue, lead value, subscription tier. A newsletter sign-up is not a purchase.

Tools: key events, event values, currency
Step 03 · Connect

Tell your ad platforms the truth

Link GA4 to Google Ads, turn on auto-tagging, import key events, share audiences, and add enhanced conversions so fewer sales go missing between the click and the checkout.

Tools: Ads link, audiences, enhanced conversions
Step 04 · Optimise

Let data, not opinions, steer budget

Value-based bidding, remarketing to high-intent users, funnel explorations and BigQuery analysis all depend on steps one to three being right.

Watch: CPA, ROAS, conversion lag, LTV
Website / App clicks · forms · buys Tag Manager events · data layer consent GA4 key events · values Google Ads bidding · audiences BigQuery raw data · forever Looker Studio reports · dashboards Website / App clicks · forms · buys Tag Manager events · data layer consent GA4 key events · values Google Ads bidding BigQuery raw data Looker reports garbage in at the top = garbage out at the bottom
Every arrow is a place data can quietly leak. A good setup checks all of them.

The usual suspects

Six GA4 setup sins I find in almost every audit

None of these throw an error. GA4 won't send you a polite email saying "your data is lying." That's what makes them dangerous.

Sin 01

The goldfish memory

Data retention left at the default 2 months. Your explorations and funnel reports quietly forget anything older, right when you need a year-on-year comparison.

FixAdmin → Data retention → 14 months. Takes ten seconds. Do it today.

Sin 02

Everything is a conversion

Page views, scrolls and button hovers all marked as key events. Your "conversion rate" looks heroic, and Smart Bidding chases scrollers instead of buyers.

FixKeep key events to real business outcomes, and give each a value.

Sin 03

Your payment gateway "drives sales"

Customers leave to pay via a gateway like bKash or SSLCommerz, come back, and GA4 credits the gateway as the source. Congratulations, your best channel is your checkout.

FixAdd payment domains to the unwanted referrals list.

Sin 04

Double-counted purchases

A hard-coded gtag and a GTM tag both fire the purchase event. Revenue looks twice as good, which is lovely right up until finance checks the bank.

FixOne tag source, one purchase event, a unique transaction_id.

Sin 05

The team is your best audience

Your staff, agency and developers visit the site all day. Without an internal-traffic filter, they are your most "engaged users," and they never buy anything.

FixDefine internal traffic by IP and activate the data filter.

Sin 06

The great "Unassigned" mystery

Messy UTMs like FB, facebook and Facebook_Ads scatter one channel across five rows, and a big chunk lands in "Unassigned."

FixA shared UTM naming sheet, lowercase only, enforced like a dress code.

The setup

What "properly set up" actually looks like

If you only fix one table's worth of things this month, make it this one. Each row directly affects how well your campaigns can optimise.

Setting → common state → what good looks like
SettingCommon stateWhat good looks like
Key eventsDefault or random events marked3–6 business outcomes, each with a value and currency
Google Ads linkNot linked, or auto-tagging offLinked, auto-tagging on, key events imported, audiences shared
Enhanced conversions"We'll do it later"On, sending hashed first-party data from forms and checkout
Consent modeBanner exists, tags ignore itConsent mode wired into GTM so tags respect choices and modeling fills gaps
Data retention2 months (default)14 months, plus BigQuery export for long-term raw data
AttributionNever looked atData-driven model, sensible lookback windows, compared against platform data weekly

Try it yourself

The 60-second GA4 health check

Tick everything that's true for your property. Be honest. Nobody's watching, except maybe your analytics, if it's set up correctly.

0/10

Proof, not theory

What clean tracking does for real campaigns

For Shohoz, Bangladesh's go-to travel-ticketing platform, the turning point wasn't a new ad format. It was measurement. Rebuilding tracking with GA4, Google Tag Manager and enhanced conversions moved optimisation from clicks to real purchases, so Target ROAS bidding could finally learn who the valuable customers were.

36.5KPurchases driven
$0.08Cost per conversion
6,521%Return on ad spend

Read the full Shohoz case study →

"Your campaigns can only be as smart as the data you feed them. GA4 is where that data is decided."

Mahdi Hasan · Performance Marketing Strategist

Key takeaways

The short version, for the people who scrolled straight here

Installed ≠ set up

The tag is the start. Key events, values, links and filters are the setup.

Automation needs good inputs

Smart Bidding optimises toward whatever you call a conversion. Choose wisely.

Privacy is part of the setup

Consent mode and enhanced conversions keep measurement honest and compliant.

Check it like a car

Audit regularly. Sites change, tags break, and nobody tells GA4.

The bottom line

Set it up once, properly, and every campaign after gets smarter

GA4 setup isn't glamorous. Nobody has ever posted a celebratory LinkedIn carousel about their unwanted-referral list. But it is the difference between campaigns that learn and campaigns that guess, and in a world where machines make most of the bidding decisions, guessing gets expensive fast.

If you've read this far and quietly scored yourself a 4 out of 10, no judgment. That's what audits are for. I help brands set up GA4, GTM and conversion tracking so their ad spend finally has something reliable to learn from. And if you're wondering what to do once the data is clean, read From Clicks to Conversions next.

Sources

  1. Google Analytics Help — Data retention
  2. Google Analytics Help — Universal Analytics has been replaced by Google Analytics 4
  3. Google Ads & Commerce Blog — Conversion modeling through Consent Mode in Google Ads
  4. Google Ads Help — About enhanced conversions