Advanced Conversion Tracking in Google Ads

In a lot of Google Ads accounts the numbers look good without that showing up in the business goals. A high ROAS can still leave barely any profit at the end of the month, and often it’s mostly returning customers driving the revenue while real new-customer growth stalls.

One reason for that can be an underdeveloped conversion tracking setup that uses far less than what’s available.

In this post I’ll walk through the advanced conversion-tracking measures and how they help you build a setup that actually pays into your real business goals.

Why is complete conversion tracking worth it?

Smart Bidding sets its bids based on the conversions you report. If that data is patchy or inaccurate, the algorithm optimizes on a distorted picture and you waste budget without noticing.

The closer your data is to reality, the better the decisions Google makes for you. That pays off concretely in three places:

• You win back conversions that get lost in the browser and give bidding more data points.
• You steer toward profit and push budget to where margin actually comes from.
• You learn what a new customer really costs and how much revenue is left after returns.

What does “advanced” mean in conversion tracking?

At the core it’s about two things: completeness and accuracy.

Completeness. Browser blockers, iOS Safari, and cookieless browsers mean a share of your purchases never lands in the conversion reporting at all. In Europe data protection adds to that: where Consent Mode applies, you only see data from users who consented, and Google models the rest. You’ll never have the complete picture. The goal is to get as close as possible.

Accuracy. By default you send revenue as the conversion value, because that’s the value sitting in the dataLayer. For good bids, though, what counts is how close that value is to the real profit and whether it accounts for returns. That means making the distinctions that matter, for example between revenue and profit or between new and returning customers.

You want more results with Google Ads?

Get a free Google Ads Reality Check: a compact video (about 10 minutes) showing where budget is slipping away in your account and what to change.

Get your Google Ads Reality Check →

Which measures should you set up?

Server-side tracking

Server-side tracking is the measure with the biggest effect on the completeness of your data. In practice it brings back around 20% more conversions, because the hits no longer fail in the browser against ad blockers, iOS Safari, and privacy settings.

On top of that you keep more control over which data goes to Google at all. And because the cookies are set from your own server, they last longer than the browser-side ones that Safari and others often cut after just a few days.

How it works: Instead of sending the data straight from the browser to Google, you put a server in between that takes in the events and passes them on.

• You set up a second GTM container, the server container, hosted at a service like Stape or in your own cloud (Google Cloud, for example).
• You set up a dedicated, inconspicuously named subdomain, something like s.yourshop.com rather than tracking., so the requests run as first-party and ad blockers don’t flag them as tracking by the name alone.
• You route the events from the web container to the server container and send the conversions to Google from there.

The setup is a one-time job, after which monthly hosting costs come on top, in the low double digits per month depending on the provider. Server-side is also the foundation that several of the following measures run better on.

Enhanced Conversions

Enhanced Conversions win back purchases that slip through standard cookie tracking, and they even work across devices. They’re one of the most accessible measures of all, because they build on data that the checkout produces anyway.

How it works: You additionally send hashed first-party data, primarily the email address. “Hashed” means one-way encrypted: Google gets no readable plain text, but can recognize two identical emails by the same hash and assign the conversion to the original click, even when the cookie is blocked.

• You enable Enhanced Conversions in the relevant conversion action in Google Ads, with Google Tag Manager as the method.
• In GTM you create a variable of type “User-provided data” that pulls the email from the checkout, ideally from the dataLayer.
• You attach that variable to your Google Ads conversion tag. Google handles the hashing automatically.

What matters is that the customer data is available on the confirmation page and consent is in place. How much it brings in the end depends on the match rate, that is, how many of the sent emails Google can assign to a logged-in account. You’ll find the full walkthrough including diagnosis in the Enhanced Conversions guide.

Cart data

By default Google only knows which product was clicked, not which products were actually bought in the end. People often click on one product, and others end up in the cart. Cart data closes that gap: you pass the actually purchased items with the conversion and can evaluate what was really sold, not just what the ad triggered.

How it works: Instead of only the total order value you report the individual purchased items, officially “Conversions with Cart Data”.

• The shop writes the purchased items into the dataLayer on the conversion event, with item_id, quantity, and price. Many shop systems deliver this via plugin, with others an adjustment in Google Tag Manager is needed.
• These item details are passed in the Google Ads conversion tag.
• The item_id has to match the IDs in the Merchant Center feed exactly, otherwise Google can’t assign the items.

This produces evaluations at the product level: you see which products are really sold through the ads and how far the clicked product differs from the purchased one.

POAS (profit instead of revenue)

POAS brings profit into bidding. Smart Bidding then bids higher on high-margin orders and holds back on thin ones. Under pure ROAS optimization a 100-euro order with 5 euros of profit counts the same as one with 50 euros of profit. As soon as your margins vary across the assortment, this is the single biggest lever on this list.

How it works: Instead of revenue you send the profit per order as the conversion value. For that you need two ingredients:

• the cost-of-goods or margin data (COGS) per product,
• and a way to replace the conversion value with the profit before it’s sent. The three common ways are a COGS feed in the Merchant Center, a server-side solution via Stape with a margin lookup, or a tool like ProfitMetrics that takes over the whole data flow.

Don’t switch from revenue to profit overnight, or you’ll throw your campaigns’ learning phase overboard. The concept and the three ways in detail are in the POAS guide.

New vs. returning customers

With this distinction you recognize which conversions come from new and which from returning customers. That matters because a “good ROAS” often rests largely on repeat buyers who would have bought anyway. For real growth, what counts is how efficiently new customers come in.

How it works: You need a piece of information that records, per order, whether the buyer is already in the system, and you pass it to Google through Google Tag Manager.

• In the shop or CRM, each order is checked for whether the buyer is already a customer, and the result is written into the dataLayer as “new” or “existing”.
• In Google Tag Manager a variable reads that value.
• Through the corresponding field in the Google Ads conversion tag it’s then passed to Google.

Once the data arrives, you can trace which revenue comes from new and which from returning customers. Building on that, you weight new customers higher in bidding or exclude returning customers from targeting via audiences, to steer the budget specifically toward new customers. For the full setup and how to steer it in the account, see the detailed guide to new vs. returning customer tracking.

Conversion adjustments (returns and cancellations)

With conversion adjustments you keep your conversion values honest by changing or deleting already-tracked conversions after the fact. That’s practical above all when you deal with a lot of returns or cancelled orders.

How it works: You report the correction to Google after the fact and assign it to the original conversion.

• The correction runs through the Google Ads API or a regular upload.
• It’s assigned to the original conversion via an order or transaction ID.
• You pull the value back entirely (full return, cancellation) or adjust it (partial return).

If 40% of the goods come back but all orders stand in the account as full revenue, Smart Bidding optimizes on revenue that never becomes real. With the adjustments it steers toward what’s actually left after returns.

For conversions to be assignable after the fact at all, you should store the Google Click ID (gclid) with every order. That’s the click parameter Google appends to the URL of every ad. With it you assign a conversion unambiguously to a click, it’s the foundation for the offline conversion import, and it additionally enables your own evaluations, for example on attribution.

FAQ: common questions about conversion tracking in Google Ads

Conclusion

Complete conversion tracking is the foundation for Google Ads being able to optimize toward your real business goals. Server-side and Enhanced Conversions win back the lost conversions, cart data shows what was actually bought, POAS brings profit into bidding, and conversion adjustments keep the values honest after returns.

You don’t have to do it all at once, and most of it is a one-time setup anyway. Start with the measure that has the biggest lever for your shop, and take on the next one once the first is running. Step by step a tracking setup emerges that reflects the reality of your business and serves as the foundation for every further optimization.

Thanks for reading this far. Good luck putting it into practice!

Work smart with Clicks in Mind