The question arrives in almost identical form every time. Someone has conversion tracking in Google Ads, they also have GA4 running, and the two numbers do not match. GA4 says forty-one enquiries last month. Google Ads says fifty-eight. Nobody can say which is right, so the monthly report quietly reports both and hopes the client picks one.
What makes this worse than an ordinary reporting annoyance is where the number ends up. Google Ads conversion tracking data does not just sit in a column. It feeds Smart Bidding. So an inflated or duplicated count is not a cosmetic problem at all, it is an instruction to spend more money chasing an action nobody has valued properly.
This piece is about deciding, not about clicking through a setup wizard. Which measurement path suits which business, and why the two platforms will never agree. Which settings actually change what gets counted, and how to tell whether the number in front of you is solid enough to bid on. The video below covers the account-level setup in about two minutes, which makes the decisions easier to follow.
Setting Up Conversion Tracking: The Account-Level Basics
Source video: Setting up Conversion Tracking Pt 1, on the Google Ads channel
PART 1 · DIAGNOSE
What a Conversion Number Is Actually For
IN BRIEFA conversion number does two different jobs, it reports to you, and it instructs the bidding system. Reporting tolerates rough numbers; bidding does not. Deciding which job a given conversion action serves is the step most accounts skip, and it sits upstream of everything else in the account build.
Google’s own documentation splits this cleanly. It describes Google Ads conversion tracking as showing what happened after someone interacted with your ad. It then explains that you can set up a separate conversion action for each type of action you care about, such as a purchase, a form, a call or an app download. It also flags the setting that decides which of those actions appear in your main reporting column.
That last part is where the money is. An account can track ten things and bid on only two. Most struggling accounts have it the other way around.
- Reporting conversions tell you what the business received. Newsletter signups, PDF downloads, scroll depth. Useful context, harmless when slightly wrong.
- Bidding conversions tell Google what to buy more of. These must be actions with real commercial value, counted once, valued honestly.
- Everything else is noise dressed as measurement, and it becomes expensive the moment it lands in the primary column.
Google publishes several different ways to track conversions, and the choice between them is a business decision before it is a technical one. Ask what one enquiry is worth in ringgit first. If nobody in the room can answer, no tracking setup will rescue the account.
BENCHMARK BRIEFING 1 OF 4
Where Your Conversion Data Can Come From
IN BRIEFFour realistic sources of Google Ads conversion tracking data exist for a Malaysian SME, and they are not interchangeable. Each carries its own setup cost, its own failure mode, and its own level of trust for bidding. That is why the choice belongs in the diagnosis stage rather than the implementation stage.
| Method | What it needs | How it usually breaks | Safe as a bidding signal |
|---|---|---|---|
| Google tag, direct on site | Tag on every page plus an event snippet | Snippet lost in a theme update | Yes |
| Google Tag Manager | A container and someone who maintains it | A trigger nobody documented | Yes |
| Imported GA4 key events | Linked accounts and auto-tagging on | Imported alongside a native duplicate | Yes, if it is the only version |
| Enhanced conversions | Hashed first-party data from your form | Fields never mapped correctly | It improves an existing signal |
Aggregated by IZI Digital Marketing from Google’s guidance on setting up conversion tracking, its documentation on creating conversions from Google Analytics events, and its overview of enhanced conversions, 2026. Failure modes are editorial observations, not Google metrics.
The fourth row is worth reading as an upgrade rather than an option. Google describes enhanced conversions as supplementing an existing conversion with hashed first-party data, such as the email address or phone number captured at your form. That lets measurement survive when cookies do not. It improves a signal you already have. It does not create one.
Not sure which of the four your account is actually running?
Plenty of accounts run two by accident, which is exactly how the same enquiry gets counted twice. See how a managed account audits the measurement layer first
PART 2 · DESIGN
Native Tag or GA4 Import: Which to Choose
IN BRIEFUse the native Google Ads tag when the action is simple and the bidding matters most. Import from GA4 when the action is already defined there and you want one definition across both platforms. Running both for the same action is the one combination to avoid, for the same reason duplicate events cause trouble on the Meta side.
Importing is not difficult. Google requires the two accounts to be linked and auto-tagging switched on, and the person doing it needs Marketer access in Analytics. What catches people out sits after the import, not during it.
DECISION BOX · NATIVE TAG OR GA4 IMPORT
| Your situation | Better choice | Why |
|---|---|---|
| One form, one thank-you page, ads are the main channel | Native tag | Fewer moving parts between the action and the bid |
| Key events already defined and trusted in GA4 | GA4 import | One definition, so the two reports argue less |
| Ecommerce with values that vary per order | Native tag with dynamic value | Value-based bidding needs the real ringgit figure |
| Several channels feeding the same enquiry form | GA4 import | You need to see what ads contributed, not just what they closed |
Verdict: Choose one source per action, mark it primary, and leave every other version secondary. Two primaries for the same enquiry is not redundancy, it is double counting with a bidding budget attached.
One detail deserves attention because it protects you by default. Google notes that conversion actions created through the Analytics interface arrive in Google Ads set as secondary, and must be changed deliberately before they influence bidding. That default is a safety net. Treat a decision to override it as a real decision.
BENCHMARK BRIEFING 2 OF 4
Why GA4 and Google Ads Never Agree
IN BRIEFThe gap between the two platforms is mostly designed, not broken. They credit different clicks, stamp conversions with different dates, and count repeat actions differently. Knowing which cause is doing the work turns a monthly argument into an ordinary diagnostic exercise.
| Cause | How often it is the answer | Which side reports higher |
|---|---|---|
| Different attribution scope | Most often | Google Ads |
| Conversion dated to the click, not the action | Very often | Varies by month boundary |
| Modelled conversions in Google Ads | Often | Google Ads |
| Counting settings differ per action | Sometimes | Either |
| Tag missing on some landing pages | Occasionally | Whichever tag is present |
Causes aggregated by IZI Digital Marketing from Google’s documentation on data discrepancies between Google Ads and Analytics and its guidance on discrepancy factors and troubleshooting, 2026. The frequency ranking is an illustrative judgment by IZI Digital Marketing, not a Google metric.
The top row explains most gaps on its own. Google states that Google Ads credits the last Google Ads click, while Analytics credits the last click across every channel. So a customer who clicks an ad, leaves, then returns through a Google search and enquires will appear as an ad conversion in one platform and an organic conversion in the other. Both are telling the truth about different questions.
The date difference compounds it. Google Ads reports the conversion against the date of the click; Analytics reports it against the date the action happened. Over a month boundary, that alone moves conversions between reporting periods.
BENCHMARK BRIEFING 3 OF 4
The Settings That Decide What Gets Counted
IN BRIEFFour settings do most of the damage when left on default: counting, conversion window, primary status, and value. Each has a sensible default for ecommerce and a misleading one for lead generation, which is where most Malaysian service businesses sit.
| Setting | Published default | What the default does | Lead-gen consideration |
|---|---|---|---|
| Counting | Every, for website actions | Counts each submission after one click | Switch to One for enquiry forms |
| Conversion window | 30 days | Ignores anything later than day 30 | Lengthen it for slow buying cycles |
| Primary or secondary | Secondary, for GA-created actions | Reports only, no bidding influence | Promote exactly one, deliberately |
| Conversion value | Often left blank | Treats every enquiry as equal | Set an average closed value in ringgit |
Defaults aggregated by IZI Digital Marketing from Google’s documentation on conversion counting options and its guidance on conversion windows, 2026. Lead-generation considerations are editorial recommendations by IZI Digital Marketing.
The counting row causes the most avoidable inflation. Google’s own example is a traveller booking three hotels and two cars, which the Every setting records as five conversions. That behaviour is correct for a hotel chain and misleading for a renovation contractor, where one person filling the form three times in an afternoon is still one enquiry.
BENCHMARK BRIEFING 4 OF 4
How Long Before the Data Can Be Trusted
IN BRIEFNew tracking is not trustworthy on day one, and neither is the bidding built on it. Google publishes a volume threshold for its default attribution model, and reaching it is what separates a real reading from an early guess, a patience requirement shared with every automated campaign type.
| Period | What to verify | Safe to change | Conclusion available |
|---|---|---|---|
| Days 1–3 | Tag fires, one action per submission | Anything broken | None |
| Week 1 | No duplicate action counting the same form | Counting and window settings | Whether the plumbing is sound |
| Weeks 2–3 | Recorded conversions match the inbox | Primary status, conversion value | Whether the count is honest |
| Weeks 4–6 | Cost per enquiry settling into a range | Budget, one step at a time | Early direction, not a verdict |
| Weeks 7–8 | Cost per closed job, not per enquiry | Bid strategy and targets | Whether the account pays for itself |
Illustrative model by IZI Digital Marketing, built on Google’s published recommendation of at least 200 conversions and 2,000 ad interactions within 30 days for data-driven attribution and its guidance that Analytics and Google Ads data require time to synchronise. A planning timeline, not measured client results.
The week two check is the one nobody enjoys and everybody needs. Open the enquiry inbox, count what actually arrived, and compare it against the conversion column for the same period. If the platform says twenty-two and the inbox holds fourteen, the tracking is wrong and no amount of bid tuning will fix it.
Want the inbox-versus-dashboard check done properly once?
It takes an afternoon and it usually changes what the account should be bidding on. Compare what a managed scope covers before you commit
PART 4 · DRIVE
Deciding Whether Your Tracking Can Be Trusted
IN BRIEFTrustworthy Google Ads conversion tracking passes three tests. One action has one source, the recorded count matches what the business actually received, and the value attached reflects what a customer is worth. Fail any one and the bidding is guessing, which is the habit IZI builds every engagement around correcting.
The three tests are worth running as a short quarterly ritual rather than a project.
- One source per action. Open the conversion actions list and look for two entries describing the same event. If both are primary, the account has been bidding on a doubled number.
- Count matches reality. Compare a full month of recorded conversions against enquiries the business can actually name. Small gaps are normal; a gap of a third is a defect.
- Value reflects worth. If every enquiry carries the same value or none at all, the bidding cannot distinguish a RM800 job from a RM40,000 one, and it will happily buy more of the cheap one.
There is a fourth question worth raising before the next quarter, and it is not technical. Google has said that from April 2026 it will accept user-provided data from website tags, Data Manager and API connections at the same time. Enhanced conversions for web and leads combine into a single setting. First-party data is becoming the measurement layer rather than an add-on to it. Businesses collecting clean contact details at the form are better placed than those that are not, and that usually comes back to how the enquiry form was built in the first place.
COMMON QUESTIONS
Google Ads and GA4 Conversion Tracking FAQ
1. Should I use the Google Ads tag or import from GA4?
Use whichever one you are prepared to bid on, and only one per action. It depends on where the definition already lives, if your key events are set up and trusted in GA4, importing keeps a single definition across both platforms. If the action is a simple form on a thank-you page, the native tag has fewer moving parts between the submission and the bid.
2. Why does GA4 show fewer conversions than Google Ads?
Usually because the two platforms credit different clicks. Google states that Google Ads attributes to the last Google Ads click, while Analytics attributes to the last click across all channels. So a visitor who returns through organic search before enquiring is counted differently by each. Modelled conversions in Google Ads and the click-date versus action-date difference explain most of the rest.
3. Will importing GA4 conversions break my bidding?
Not by itself, because of a helpful default. Google sets conversion actions created through the Analytics interface as secondary, meaning they report without influencing bidding until someone changes that. The risk appears when an imported action is promoted to primary while a native tag is still counting the same event.
4. What conversion window suits a Malaysian service business?
Longer than the default, in most cases. Google’s default window is 30 days and applies changes only going forward. For a considered purchase, renovation, medical, education, B2B equipment, a decision commonly lands after that, so a 30-day window quietly discards conversions your ads genuinely earned.
5. Do I still need conversion tracking if I only get WhatsApp enquiries?
Yes, though it needs more care. The click to open WhatsApp is trackable, but a click is not an enquiry, plenty of people open the app and never send a message. Count the click if you must, keep it secondary, and find a way to record which conversations became real enquiries before you let anything bid on it.
Not sure whether the number your account bids on is real?
A Blueprint session reads your conversion actions, your GA4 setup and your actual enquiry volume side by side, then tells you plainly which number to trust and which settings to change. You leave with the decision made, not a list of options.