Is your web agency telling you Google Ads is doing brilliantly because you can see hundreds or thousands of conversions? Before you believe it, check one fundamental thing: are those conversions really paid orders, or are they simply add-to-cart events?
In the case we look at here, an eCommerce campaign shows 978.06 conversions at a cost of roughly €911.57. At a glance it looks like an excellent result: less than €1 per conversion. But open the goal breakdown and the problem appears: around 874 conversions are "Add to cart", while real purchases are around 103.
Translated: more than 89% of the declared conversions are not sales. They are carts, and a cart doesn't pay salaries, doesn't generate margin, doesn't empty the warehouse and doesn't really measure the profitability of an eCommerce.
The problem: calling events that aren't purchases "conversions"
To be clear: tracking add-to-cart is not wrong. It is genuinely useful for understanding user behaviour in the funnel. Google Analytics 4 clearly distinguishes eCommerce events such as add_to_cart and purchase: the first measures interest, the second measures the purchase.
The problem starts when a web agency or a Google Ads consultant uses add-to-cart as the primary conversion, then presents the total conversion count to the client as if those were purchases. That distorts how results are perceived and can make a campaign look profitable when in reality it is generating plenty of intent and very few orders.
According to the official Google Ads documentation, primary conversion actions are included in the "Conversions" column and can be used for bid optimisation. Secondary actions, on the other hand, are for observation and reporting. For an eCommerce, as a rule, the primary goal has to be the purchase, not the cart.
Why it is serious in Performance Max campaigns
With Performance Max campaigns, Google uses automation, machine learning and conversion signals to decide where to show the ads, who to show them to and how much to bid. Google itself recommends choosing conversion goals that genuinely matter to the business.
If the campaign learns that the result to chase is "add to cart", it will tend to find people inclined to add products to a cart. Not necessarily people inclined to complete the order, pay and become customers.
That is the most dangerous part: this isn't just an inflated report. It is a configuration that can damage the campaign's learning. Instead of optimising towards revenue, margin and purchases, Performance Max can end up optimising towards micro-actions that are far easier to obtain.
How an aggregated report can inflate the perception of results
The pattern is simple:
- set "Add to cart" as the primary conversion action;
- leave "Purchase" bundled with other events under the same goal logic;
- show the client the total "Conversions" column;
- talk about an extremely low cost per conversion;
- hide or play down the number that actually matters: how many orders were generated.
In the screenshots, the aggregated report shows around 978 conversions. But the breakdown clearly separates 873.98 carts from 103.08 purchases. They are two completely different metrics.
Which signals to check in your Google Ads agency's reports
If you run an online shop and suspect your agency is inflating the results, check these signals:
- They always talk about conversions, never about orders. Ask how many paid sales came from Google Ads.
- The cost per conversion looks too good. If you are paying less than €1 per conversion but orders aren't growing, you may be looking at carts, not purchases.
- You don't have admin access to Google Ads and GA4. If you can't check the goals yourself, you depend on the agency's version of events.
- In the Goals panel, cart, begin checkout and purchase all appear as primary actions. For analysis that's fine; for bidding and commercial reporting it can be a disaster.
- Performance Max is spending but revenue isn't following. That is the classic symptom of campaigns optimised on weak signals.
- The monthly report doesn't separate micro-conversions from macro-conversions. A serious report separates carts, checkouts, purchases, order value and ROAS.
The correct configuration for an eCommerce
For most eCommerce businesses, a clean configuration should draw these distinctions:
- Purchase: primary conversion, used for bidding and optimisation.
- Purchase value: the real order amount, essential for ROAS and value-based campaigns.
- Add to cart / add_to_cart: secondary event or observation metric, useful for diagnosing the funnel.
- Begin checkout: secondary event, useful for understanding abandonment and UX problems.
- Leads, calls, store visits: only to be used if they are genuinely goals consistent with the campaign.
There are technical exceptions, for example accounts with very few sales or temporary learning strategies. But they have to be declared, documented and measured. They cannot become the standing way to make campaigns look like winners when they are not generating orders.
Google Partner status is not a guarantee
Many eCommerce owners trust an agency because it displays badges, certifications or labels such as "Google Partner". Those elements can indicate skills and requirements met, but they do not replace checking the data. Even an account run by certified professionals can be badly configured, badly read or told even worse.
The badge is not the point. The point is: is the campaign bringing in profitable orders? If the answer gets replaced by "you got loads of conversions", ask immediately for the breakdown by conversion action.
How to protect your budget and check the agency's work
You don't need to be technical to run the first checks. You need to insist on transparency.
- Ask for admin access, or at least read access, to Google Ads, GA4 and Google Tag Manager.
- Open Google Ads and go to Goals → Conversions → Summary.
- Check which actions are set as primary conversion actions.
- In the campaign report, segment the Conversions column by conversion action.
- Compare Google Ads conversions, real orders in your management system and revenue in the CMS.
- Ask for a report with purchases, purchase value, cost, ROAS, estimated margin and new customers.
- If the agency dodges the question or changes the subject, that is data in itself.
For a complete check you can start from our 20 checks for a Google Ads audit, compare the figures with our guide on misleading ROAS and see why Performance Max shouldn't optimise on carts when the real goal is orders.
Frequently asked questions
Is my web agency ripping me off if it uses "Add to cart" as a conversion?
Not always. Using the cart as an analysis event is normal. It becomes a problem when it is used as the primary conversion to judge the campaign's success or, worse, when it is presented as though it were a purchase.
How do I tell whether Google Ads conversions are real orders?
Go into Google Ads, open the campaign report and use the segment by conversion action. You need to see how many conversions come from "purchase" or an equivalent event, and how many come from "add_to_cart", "begin_checkout" or other micro-actions.
Can Performance Max get worse if it optimises on carts?
Yes. If the main signal is the cart, the campaign can learn to look for users who create carts, not users who buy. For an eCommerce, that can degrade the quality of the learning and push budget towards less profitable traffic.
What is the correct primary conversion for an eCommerce?
Usually the completed purchase, with the order value correctly passed to Google Ads or GA4. Carts, checkouts and product views are useful, but they should stay as diagnostic metrics except in very specific, temporary strategies.
Can a Google Partner agency make these mistakes?
Yes. The Google Partner badge is not an absolute guarantee about how a particular account is configured. You should always check what is being tracked, what is being used for bidding and what is being presented in the report.
What should I ask my Ads agency every month?
Ask for the number of orders, order value, spend, ROAS, cost per purchase, estimated margin and a breakdown of conversions by individual action. If the report contains only aggregated "conversions", the most important figure is missing.
Can a badly optimised account be recovered?
Yes, but you need to put the tracking back in order, set purchases as the primary conversion, verify the values, check GA4 and Google Tag Manager, then give the campaigns a fresh learning period with the correct goals.
Want to know whether your Google Ads agency is reporting real conversions or carts dressed up as sales? We can run a technical audit of your Google Ads, GA4 and Tag Manager setup and tell you exactly what the campaign is learning: orders or illusions.
Request an independent audit of your Google Ads conversions
Useful sources: Google Ads documentation on primary and secondary conversion actions, the Google Ads guide to conversion goals, Google's best practices for Performance Max and the GA4 documentation on eCommerce events.