> ## Content Index
> Fetch the complete content index at: https://cascader.io/blog/llms.txt
> Use this file to discover other available public pages before exploring further.

# My new Shopping campaign has clicks but no carts—would switching to Performance Max help?
- URL: https://cascader.io/blog/shopping-clicks-no-carts-switch-performance-max/
- Published: 2026-09-23T20:00:35.000Z
- Updated: 2026-10-07T19:08:09.000Z
- Description: Before switching a new Shopping campaign to Performance Max, check what the missing carts mean, what each basket can contribute, and whether the next observation is worth paying for.
- Author: Rick Hecker
- Tags: Cascader Answers, #Import 2026-09-20 14:50

Switching to Performance Max might help you find buyers, but **clicks with no recorded carts do not yet tell you that campaign type is the problem**. Before funding a switch, establish whether a real cart action is recorded, what those visitors were offered, and how much a completed basket can contribute towards advertising. If another day's spend would hurt, request a pause now while you do this.

Maximize Clicks is doing the job its name describes: trying to buy as many clicks as possible within your budget. It does not select bids to maximise purchases. Performance Max uses conversion or conversion-value bidding and reaches more of Google's inventory, using the same Merchant Center product data for Shopping ads. Those are meaningful changes, but neither fixes a broken cart button or makes an unprofitable basket profitable. [Google's bidding explanation](https://support.google.com/google-ads/answer/6268626?hl=en&ref=cascader.io), [Shopping and Performance Max comparison](https://support.google.com/google-ads/answer/2454022?hl=en&ref=cascader.io).

Suppose you run a small UK jewellery shop on Shopify. Meta has been consuming the profit, so you try Standard Shopping at £20 a day. Three days in, one day shows about £40 spent, 97 clicks and no recorded carts. That is roughly 41p per click. It is one day's observation, not 97 clicks spread across the entire three-day trial—and “recorded” is doing important work.

## First, make one cart visible

A checkout can accept payments while the advertising system receives no useful event. Conversely, an event can fire even when the wrong product, price or delivery charge reaches the basket. Check the shopping action and its record together.

Choose a product that actually received clicks. Open its landing URL directly on a phone, without clicking your own ad. Select the advertised variant, add it to the basket and proceed far enough to see the delivered price. Keep the time, product, device and consent choice beside the result.

If your measurement uses a Shopify app pixel, go to **Settings → Customer events → the app pixel's menu → Test**. Shopify Pixel Helper shows received events as you navigate. If it is awaiting consent, use the store's consent controls for your test. A green event means the pixel's event-handling code succeeded; it does not by itself prove Google received the right conversion. [Shopify's app-pixel test instructions](https://help.shopify.com/en/manual/promoting-marketing/pixels/app-pixels?ref=cascader.io).

For the Google & YouTube app, also check which Google destination each Shopify event is sent to. In [Google Tag Assistant](https://support.google.com/tagmanager/answer/10039345?hl=en&ref=cascader.io), use **Add domain → your store URL → Connect** and review the tag activity as you navigate. Follow a test purchase through to the expected purchase action and value, checking for duplicate measurement.

A test made without an ad click is a test of the event path, not evidence that the campaign acquired a sale. If you use another implementation, have its installer demonstrate this same chain; do not add a second integration simply to make a number appear. [Google's Shopify conversion setup and verification](https://support.google.com/google-ads/answer/13494537?hl=en&ref=cascader.io).

In the Google & YouTube app integration, Shopify's `product_added_to_cart` maps to Google's `add_to_cart`, and `checkout_completed` maps to `purchase`. Check the item and currency as well as the event name. The default purchase value excludes shipping and tax and accounts for discounts; it is not your contribution after costs. [Google's event mapping](https://developers.google.com/tag-platform/gtagjs/reference/shopify-event-parameters?ref=cascader.io).

The result changes what to do next:

- **The item does not enter the basket, or the advertised offer cannot be bought.** Repair that specific failure before paying for more visits.
- **The basket works but its expected event fails.** Repair measurement. Today's test cannot recover yesterday's missing cart history.
- **The action and its record both work.** You have verified this route under these conditions. Now inspect actual visitor progression; one successful test cannot establish that all visitors had the same experience.

If the expected pixel is absent or you cannot verify its destination, save the test details for whoever owns the setup. Leave the campaign paused if you cannot afford to buy traffic whose outcome you cannot interpret. You have a useful next job even without a diagnosis of the old clicks.

## Read the funnel for the campaign you are judging

In Shopify's **Growth** reporting, open the channel report, choose **Google**, then the particular campaign and its campaign attribution report. Use the dates of the clicks you are investigating. The campaign report includes an online-store conversion funnel; its attribution depends on UTM parameters and connected-app data. Shopify says marketing metrics can take up to 24 hours to update. [Shopify's marketing reports](https://help.shopify.com/en/manual/promoting-marketing/analyze-marketing/marketing-performance?ref=cascader.io).

Read the campaign's sessions, carts, checkouts and orders together. A Google-wide row can mix paid and unpaid traffic, and an all-store cart count cannot describe one Shopping campaign. Do not divide that count by 97 Google Ads clicks: sessions and clicks are different units.

If campaign identification was never retained, you cannot manufacture it from today's settings. Use the broader report only for a broader question—for example, whether *any* visitors are reaching carts—and verify campaign identification before a future paid observation. That future record will not retrospectively identify the old visitors.

Now suppose the identifiable Shopping visitors still show no carts, while the test route works. Inspect the products and searches that spent the money. In Google Ads, filter to the campaign and period, inspect its product performance, then open **Insights and reports → Search terms**. Look for a concrete mismatch: an ad for plated jewellery appearing for a product specification you do not sell, or clicks landing on an unavailable variant. Correct the feed or exclude a clearly unsuitable search. A relevant search with zero sales is not automatically unsuitable, and the search-term report omits some low-volume queries. [Google's search-term reporting guidance](https://support.google.com/google-ads/answer/2472708?hl=en&ref=cascader.io).

If those checks uncover no mismatch, the remaining uncertainty is real: visitors may not want this offer, may need more time, or may simply be too few to establish a pattern. Google documents that purchases can arrive after the click, making recent performance look worse than a mature period. That warrants revisiting the existing dates; it does not require buying more clicks while you wait. [Google's conversion-delay explanation](https://support.google.com/google-ads/answer/6239119?hl=en&ref=cascader.io).

## Put the actual basket beside the 41p click

A working route tells you that someone *can* buy. The next question is whether enough purchases at the available margin could pay for the traffic.

Take an illustrative offer: £30 for one piece, £3 delivery below a £35 threshold, and a two-piece bundle discounted from £60 to £55 with free delivery. The delivery service may be perfectly good. The decision now depends on the delivered offer the visitor sees and the money the seller keeps.

Use your selling prices after discounts, delivery charges collected, supplier invoices and fulfilment costs to cost each plausible basket. Subtract product cost, packaging, postage, payment fees, applicable tax owed and an allowance for returns or refunds. What remains before advertising is **contribution per order**. The following costs are invented to demonstrate the calculation, not suggested jewellery margins:

| Per completed order                           | One £30 piece  | Two-piece £55 bundle |
| --------------------------------------------- | -------------- | -------------------- |
| Customer pays, including delivery             | £33            | £55                  |
| All the non-advertising deductions above      | £23            | £35                  |
| Contribution available before advertising     | **£10**        | **£20**              |
| Purchase rate needed at roughly 41p per click | **About 4.1%** | **About 2.1%**       |

The last row comes from:

**Break-even purchases per click = average cost per click ÷ contribution per order.**

For the bundle, £0.412 ÷ £20 is about 0.0206: just over two purchases per 100 clicks. At that rate, the contribution approximately covers the advertising. It leaves nothing from those orders for fixed overhead or additional profit, so the advertising cost you can afford per order should be lower than that order's full contribution.

The bundle is easier to fund *in this illustration*, but that is no reason to assume visitors will choose it. If the paid visitors want a single piece, using the bundle's £20 contribution would halve the purchase rate you think you need. An average product price of £55 does not establish an average basket, let alone its contribution.

This also shows why “offer free shipping” is incomplete advice. On the single piece, absorbing another £3 of delivery would reduce the illustrative contribution from £10 to £7 unless something else changes. At the same CPC, the required purchase rate rises to about **5.9%**. Free delivery could still help if enough more visitors buy; the arithmetic tells you what the improvement must pay for.

Before changing the offer, watch someone unfamiliar with the shop use that product page. Ask them to find what they would pay delivered and choose between the single piece and bundle. If they cannot find the charge, misunderstand the bundle, or cannot select a variant, you have something observable to fix. If they understand and decline, ask what prevented the choice. This inexpensive reading can expose friction or objections; it cannot supply a campaign conversion rate.

With no order history, cost the likely single-item and bundle baskets separately and use a cautious allowance where costs remain uncertain. If you cannot establish positive contribution, stop paid acquisition until you can. If contribution is positive but the purchase rate is unknown, you have a conditional opportunity—not a promise that either Google campaign will outperform Meta.

## Choose the change that addresses what you found

**Keep Standard Shopping for a limited observation** if the path is measured, the offered products have plausible economics, and your immediate question is how visitors respond to those products. Keep the product set, offer and measurement consistent during that observation. A CPC limit with Maximize Clicks, or Manual CPC for product-group control, can constrain bidding; neither guarantees suitable buyers or profit. Bid adjustments can apply above a Maximize Clicks limit. [Google's CPC-limit guidance](https://support.google.com/google-ads/answer/6268626?hl=en&ref=cascader.io).

**Change bidding without changing campaign type** if the important change is to pursue order value instead of click volume. Google's current documentation lists Maximize conversion value and Target ROAS for Shopping. Maximize conversion value requires transaction-specific conversion tracking, uses past performance, and aims to spend the budget. It is a capability to consider once trustworthy order values are arriving, not a reason to spend through an unverified zero. [Shopping conversion goals](https://support.google.com/google-ads/answer/9546836?hl=en&ref=cascader.io), [value-bidding prerequisites](https://support.google.com/google-ads/answer/7684216?hl=en&ref=cascader.io).

**Choose Performance Max** when you deliberately want conversion/value bidding *and* broader reach across Google's channels, with a verified purchase goal, useful assets and money you can risk on learning whether that reach pays. It may find a better audience; this early zero cannot establish that it will. Keep cart events for diagnosis rather than promoting them to the purchase objective merely because they are more plentiful. Google's primary actions guide bidding when their goal is selected; even secondary actions can be used for bidding inside a custom goal. [Primary and secondary conversion actions](https://support.google.com/google-ads/answer/11461796?hl=en&ref=cascader.io).

If you already launched PMax, record the start date and its spend separately. A subsequent sale would be welcome, but a before-and-after change would not prove PMax beat Shopping. Google offers a formal Shopping-versus-PMax experiment; splitting this tiny amount of affordable traffic is not the necessary next step. [Google's experiment setup](https://support.google.com/google-ads/answer/13825980?hl=en-EN&ref=cascader.io).

## Decide what uncertainty is worth buying

A £20 **average** daily budget normally permits a £40 billed day. For most campaigns, the monthly limit is 30.4 times an unchanged average daily budget—£608 for a full month here. These are platform spending rules, not an affordable research budget. [Google's spending limits](https://support.google.com/google-ads/answer/10486637?hl=en&ref=cascader.io).

Set an allowance for further loss with no new order, including costs that may still accrue while a pause takes effect. Choose the observation that would justify reviewing the decision. Suppose you set a £30 allowance. At the previous CPC, £30 corresponds to roughly 73 clicks, which may still leave you undecided. That comparison gives the scale of the expense; it is not a promise to buy exactly 73 clicks or stop at £30.

Google says ads can continue serving for several hours after you pause a campaign, and costs incurred before they actually stop remain payable. This is separate from old costs appearing late in a report or a later bill. A current spend report and a saved pause do not prove serving has ended; Google gives no guaranteed reserve amount or exact completion time here. [Google's charges FAQ](https://support.google.com/google-ads/answer/16304798?hl=en&ref=cascader.io).

**If £30 is a hard maximum you cannot risk exceeding,** do not fund another observation on the assumption that a pause can enforce it. **Choosing no new paid observation is valid.** If a campaign is already serving, request its pause promptly; this cannot erase incurred costs or guarantee zero further cost while serving ceases. If it is already paused, leave it paused. [How to pause a campaign](https://support.google.com/google-ads/answer/2404259?hl=en&ref=cascader.io).

**If you can afford uncertain remaining exposure and choose to proceed,** leave part of the allowance unused for it, but do not assume any chosen headroom guarantees the ceiling. Set a budget you can tolerate even on an overdelivery day and request a pause well before you expect to exhaust the allowance; **do not wait for reported spend to reach £30**.

**Request a pause for review sooner if a verified order arrives**, so you can cost the basket actually bought. Neither a cart nor an order resets the allowance.

Review later outcomes and existing records with the campaign left paused, without starting another paid cohort. A cart count shows progression; recorded item details, if available, show the basket.

For a tiny shop, the useful outcome may be a repaired event, a clearer delivered offer, a basket you can afford to advertise—or a decision to keep both campaigns paused. You do not need to prove that Shopping has failed before deciding that the next paid answer costs too much.