If you’re running Google Ads and you’re not completely certain your google ads conversion tracking is set up correctly, there’s a decent chance you’re making decisions based on numbers that don’t mean what you think they mean. Not a little off. Potentially completely backwards. We’ve audited enough accounts to say that with some confidence, and the pattern shows up across industries, budgets, and business types. Small accounts, large accounts, accounts managed by solo business owners, accounts managed by other agencies. The same fundamental tracking problems appear again and again.
The frustrating part is that the platform will not tell you something is wrong. Google Ads will happily report hundreds of conversions, show you a cost-per-conversion you feel okay about, and let you optimize campaigns toward those numbers, even if those conversions are phantom clicks, double-counted form submissions, or people who bounced in four seconds. The dashboard looks fine. The campaigns look like they’re working. You keep spending.
So before we get into the mechanics of how to actually set this up, it’s worth sitting with that for a second. The reason tracking setup matters isn’t just technical hygiene. It’s that every budget decision you make flows downstream from what the platform thinks is working. Garbage in, garbage out, as they say, except in this case the garbage looks clean and comes with a nice graph.
What a Conversion Actually Is, and Why People Get This Wrong
Most people have a loose definition of conversion in their head. Someone did something good. They filled out a form. They called. They bought something. That’s accurate enough as a concept, but the platform requires you to be specific in ways that force you to make real decisions, and a lot of businesses skip those decisions and just accept whatever the default setup produces.
A conversion action is a specific, defined event that Google Ads tracks and counts. You get to decide what counts. The problem is that most businesses either define too many things as conversions or define the wrong things. We’ve seen accounts where someone counted every single page visit as a conversion event. We’ve seen accounts where “time on site over two minutes” was being treated as a primary conversion. Those numbers get fed into Smart Bidding algorithms, and the algorithm starts optimizing for whatever behavior you’ve told it to value, whether or not that behavior has anything to do with actual revenue.
Before you touch a single setting, write down the answer to this question: what does someone have to do for you to consider that ad spend justified? Not “what would be nice to know about.” What actually matters. For most service businesses, that’s a phone call or a completed contact form from someone who isn’t already a customer. For e-commerce, it’s a purchase. Everything else is secondary, and in most accounts it should be flagged as such rather than mixed in with your primary conversion data.
Primary vs. Secondary Conversions and Why the Distinction Matters More Than People Think
Google Ads lets you designate conversion actions as either primary or secondary. Primary conversions get included in your “Conversions” column, the one Smart Bidding strategies use to optimize. Secondary conversions show up in a separate column called “All Conversions” and are visible but not baked into the algorithm’s decisions. Most accounts we look at have not made this distinction deliberately. Everything is primary, or nothing is organized at all.
Here’s why that creates problems. If you’re using Target CPA or Maximize Conversions bidding, the algorithm is actively learning from your primary conversion data and adjusting bids in real time to get you more of whatever those conversions are. If your primary conversions include both phone calls from new customers and newsletter signups from existing clients, the algorithm can’t tell the difference. It will pursue whatever is easiest to get. Newsletter signups from existing clients are much easier to get than phone calls from strangers who’ve never heard of you. The algorithm will drift toward easy wins, your call volume drops, and you wonder why the campaigns seem to be generating less business even though the conversion numbers look fine.
The fix is simple but requires a real decision: set your actual revenue-driving actions as primary conversions only. Phone calls over a meaningful duration (more on that in a moment), completed lead forms from new prospects, purchases, booking confirmations. Everything else, page engagements, soft clicks, scroll depth, newsletter signups, gets marked secondary. You can still see the data. It’s useful. It just doesn’t drive your bidding strategy.
Setting Up Phone Call Tracking Without Stepping on Yourself
Phone calls are where tracking gets messy fast, especially for service businesses where calls are often the highest-value lead type. There are actually three different ways to track calls through Google Ads, and using the wrong one, or using more than one simultaneously without understanding what each counts, is a very common source of inflated conversion numbers.
The first method tracks calls from your ads directly, meaning someone sees your ad on a mobile device and taps the phone number without ever visiting your website. Google routes the call through a forwarding number and records it as a conversion. The second method uses a Google forwarding number on your website, placed dynamically so that visitors arriving from a Google Ads click see a different number than organic visitors. The third method involves importing call data from a third-party call tracking platform like CallRail into Google Ads.
Where people get into trouble is running two or three of these at once. Someone clicks your ad, lands on your site, calls the number they see, and that call gets recorded as a conversion from the website forwarding number and potentially from the ads click-to-call as well, depending on how things are wired. One phone call, counted twice. Multiply that across a month and your cost-per-conversion looks better than it is. You think your campaigns are more efficient than they actually are. You scale up. Reality doesn’t scale with you.
Pick one primary method for tracking calls and make sure the others are either disabled or set to secondary. For most businesses using a CRM or call tracking software, importing from that system gives you the cleanest data because you can apply filters, flag the calls that were too short to be real leads, and only count what actually matters. If you’re not using a third-party call tool, the Google forwarding number on the website is generally more reliable than relying solely on click-to-call tracking from ads.
And set a call duration threshold. The default in Google Ads is 60 seconds, which is too short in most industries. A 60-second call could be someone who dialed the wrong number and stayed on the line briefly. For service businesses, we typically recommend setting the threshold at 90 seconds minimum, and often higher depending on how your intake process works. A legitimate lead for an HVAC company or a plumber or a roofing contractor usually involves at least two to three minutes of conversation. Count the short ones as secondary if you want to see the volume data, but keep them out of your primary conversion set.
Form Submission Tracking and the Tag That Fires Every Time Someone Refreshes the Page
Form submission tracking sounds simple. Someone submits a form, a conversion fires. But the implementation details create a lot of ways for this to go sideways, and the most common one involves a thank-you page that isn’t quite set up the way you think it is.
The cleanest way to track form submissions is to redirect users to a dedicated thank-you page after they submit the form, then place your conversion tag or Google Tag Manager trigger on that thank-you page. Simple, reliable, hard to accidentally double-count. The problem is that a lot of websites either don’t have a dedicated thank-you page, show an inline success message on the same page, or have a thank-you page that’s accessible directly without going through the form. In the last case, anyone who bookmarks the URL and revisits it later fires a conversion. Anyone who refreshes the page fires another one.
If your site uses inline form confirmation messages rather than a redirect, you’ll need to use an event-based trigger instead of a page view trigger. This means detecting when the form actually submits and firing the tag at that moment, not when the page loads. In Google Tag Manager, this typically involves a Form Submission trigger or a custom event pushed from your form plugin. It’s a bit more setup but it’s the right approach for inline confirmations.
One thing worth checking regardless of how your form is set up: open an incognito browser, submit your own contact form as a test, and then look at the real-time conversion report in Google Ads or the DebugView in GA4. You should see the conversion fire exactly once. If you see it fire more than once, or not at all, you’ve found your problem before it distorts a month of data.
Google Tag Manager Is Not Optional at This Point
Some people are still placing Google Ads conversion tags directly in their website’s code, usually somewhere in the header or footer, without using Google Tag Manager. This works, technically, but it creates a fragile system where any change to your tracking requires a developer, any mistake in the code affects the whole site, and auditing what’s actually firing requires someone to dig through raw HTML. Google Tag Manager makes all of this manageable by giving you a single container that holds all your tags and lets you control what fires, when, and under what conditions, without touching the site’s codebase for every change.
If you’re not using Google Tag Manager yet, the setup involves placing a single GTM container snippet in your site’s code, and then everything else, Google Ads tags, GA4, conversion events, remarketing tags, lives inside the GTM interface. You can test before publishing, roll back if something breaks, and build triggers that are responsive to actual user behavior rather than just page loads.
The other reason GTM matters is that it gives you access to the Tag Assistant debugging tool and the GTM preview mode, both of which let you walk through a user session and see exactly which tags are firing and why. When a client tells us their conversions look off and they can’t figure out why, nine times out of ten the answer is in preview mode within about ten minutes.
The Smart Bidding Problem Nobody Talks About
There’s a broader issue here that’s worth naming directly. Google’s Smart Bidding strategies, Target CPA, Target ROAS, Maximize Conversions, Maximize Conversion Value, are genuinely powerful when they’re working from accurate data. Google’s own research suggests that advertisers using Smart Bidding with clean conversion data see meaningful improvements in cost efficiency. But those algorithms need a minimum volume of conversions to function properly, and they need those conversions to represent real outcomes.
The official recommendation from Google is at least 30 conversions per month per campaign for Target CPA to work reliably, and ideally 50 or more. A lot of smaller advertisers are running Smart Bidding with 8 or 12 conversions a month, half of which are inflated by tracking errors, and wondering why performance is inconsistent. The algorithm is learning from noise. It’s not being dumb. It’s doing exactly what it was designed to do. The inputs are just wrong.
This is why tracking accuracy isn’t a “nice to have” or something you clean up later. If you’re running Smart Bidding, your entire campaign performance is downstream of whether the data you’re feeding the algorithm reflects reality. Fix the tracking first. Then optimize bidding. Doing it the other way around is rearranging furniture in a house with a bad foundation.
Connecting Google Ads to GA4 and Why Both Matter
Google Ads conversion tracking and Google Analytics 4 are separate systems, and a lot of businesses treat them as interchangeable when they’re not. You can import GA4 goals into Google Ads and use them as conversion actions, or you can set up native Google Ads conversion actions directly. Both approaches have their place, but mixing them without understanding the differences creates reporting inconsistencies that are hard to diagnose.
The main practical difference is attribution. Google Ads conversion tracking uses data-driven attribution by default (or last click in some configurations), and it only counts conversions that came from someone who clicked one of your ads. GA4 reports all conversions across all channels and uses its own attribution models. When you import a GA4 conversion into Google Ads, you’re bringing in that event, but how it gets counted and attributed can differ from what you’d see in native Google Ads tracking, especially for assisted conversions.
The cleanest setup for most businesses is to run both in parallel. Native Google Ads conversion actions for the things that drive your bidding strategy, and GA4 connected to your Google Ads account via the linked property for broader visibility into how ads-driven traffic behaves on your site after the click. GA4 is particularly useful for understanding what happens between the click and the conversion. Are people from your ads visiting multiple pages before converting? Are they converting on the first visit or coming back later? That context is hard to get from Google Ads reporting alone.
Remarketing Tags and Audience Building
While we’re talking about tags and tracking, it’s worth mentioning the remarketing tag because it’s often either missing entirely or placed incorrectly. The remarketing tag (or the Google Ads remarketing event snippet, if you’re using the global site tag approach) is what allows you to build audiences from your website visitors and target them with ads later. If it’s not on every page of your site, you’re building incomplete audiences. If it’s not firing at all, you’re locked out of remarketing entirely.
For most businesses, setting up a few core remarketing audiences is worth doing early, even if you’re not actively running remarketing campaigns yet. Past visitors, people who reached the contact page but didn’t submit the form, people who spent more than a certain amount of time on the site. These audiences build over time, and you can’t retroactively populate them. When you’re ready to run a remarketing campaign, you want the audiences to already be there with real data in them, not waiting 90 days for them to fill up.
A Simple Audit You Can Do Right Now
If you’re not sure whether your current tracking setup is accurate, here’s a quick way to pressure-test it without needing a developer on call. First, open your Google Ads account and navigate to Tools and Settings, then Conversions. Look at every conversion action listed. For each one, ask yourself: is this something that actually indicates a real lead or sale? Is this marked as primary or secondary? Does the count look reasonable relative to the traffic you’re sending?
Second, check the “Conversion Source” column. You should be able to see whether each conversion is tracked via website tag, phone calls, imported from GA4, or another source. If you see the same type of event showing up from multiple sources, there’s a good chance you’re double-counting.
Third, look at the “Tag Status” indicator for each website-based conversion. Google checks whether it’s seen your tag fire recently and will flag it as “Active”, “Inactive”, or “No recent conversions.” An inactive tag is a broken tag. No recent conversions can mean the tag is fine but nobody has converted, or it can mean nobody has triggered the tag because the setup is wrong. Worth investigating either way.
Fourth, use the Google Tag Assistant Chrome extension to verify what’s firing on your actual pages. Install it, navigate to your site, and look at what tags are present and whether they’re firing cleanly. This is the fastest way to spot tags that are present but misconfigured.
It’s a 20-minute exercise that can surface problems that have been distorting your data for months. We’ve done this with new clients and found situations where a campaign that looked profitable on paper was actually running at a loss once the phantom conversions were stripped out. That’s not a fun conversation to have, but it’s a much better conversation than continuing to scale a campaign built on bad math.
When to Ask for Help
There’s a version of this where you roll up your sleeves, work through the steps, and end up with a clean, accurate tracking setup that you fully understand and can maintain. That version is achievable, especially if you’re reasonably comfortable in Google Ads and Google Tag Manager.
There’s also a version where the setup is more complicated because of your website platform, your CRM, your call tracking tool, or the way your forms work, and getting everything to fire correctly requires more than an afternoon of troubleshooting. If you’ve made several passes at your tracking setup and you’re still not confident the numbers are right, that’s a reasonable point to get outside eyes on it. Bad tracking data is expensive in a very specific way: it’s invisible. You don’t know what you’re losing. You just keep spending and wondering why results feel slightly off from what the reports are saying.
At Lost & Found Marketing, one of the first things we do with any new account is a full tracking audit before we touch campaigns, bids, or budgets. Not because we like audits, but because running ads without knowing what you’re actually measuring is a fast way to burn through budget and blame the wrong things when results don’t come. The tracking tells you what’s real. Everything else is a guess dressed up as strategy.
If you work in a specific industry and want to understand how tracking fits into a broader paid search approach, take a look at what we do across the industries we serve. The principles of clean conversion tracking apply everywhere, but the specific conversion actions that matter most, and the thresholds that indicate a real lead versus a noise event, vary quite a bit depending on your business model and sales process.
The bottom line is that google ads conversion tracking is not a setup task you do once and forget. It needs to be checked periodically, especially after website updates, form changes, CRM migrations, or any time you add or remove campaign types. Tags break. Redirects change. Thank-you pages get redesigned. Any of those things can quietly take down your tracking and leave you flying blind for weeks before you notice the numbers look odd.
Build the habit of checking tag status monthly. Test your forms in incognito. Look at the Conversions column with a critical eye whenever something feels off. The platform rewards advertisers who know what they’re measuring, and it will cheerfully mislead the ones who don’t. That sounds harsh, but it’s just how automated systems work. They optimize toward the signal you give them. Give them a clean signal and they perform. Give them noise and they’ll optimize toward the noise with impressive efficiency.
You’ve put real money into these campaigns. The least you can do is make sure you actually know what’s happening with it.
Take your marketing to the next level with Lost & Found Marketing. We work with businesses across industries to clean up their tracking, build campaigns on solid data, and stop wasting budget on numbers that don’t reflect reality. Book a call with us today and let’s take a real look at what your Google Ads account is actually telling you.