The order explorer explained: what your ads actually sold

What is the order explorer for?

The order explorer answers three questions:

1. How well do we know where the orders came from?
2. Which ads actually got the credit for the sales, and did they pay off?
3. Did customers buy the product the ad showed, or something else?

Google and Meta answer the same questions in their own tools, but they count their own way and often give themselves more credit than they deserve. The order explorer only counts what Promoosh actually saw: an ad click from the same visitor before the purchase.

One example that explains most of it

Kari sees a Meta ad for an expensive diagnostic tool on Monday and clicks. She looks around but does not buy. On Wednesday she searches Google, clicks a Google ad and buys a much cheaper tool.

Under "Did customers buy what the ad showed?" the Meta ad for the expensive tool will show an order where the customer did not buy the same product, but the cheaper one. Meta itself will often report a sale of the expensive product.

The choices at the top

Period decides which orders are included.

Attribution window decides how old an ad click can be and still get the credit. With seven days, only clicks from the week before the purchase count. A shorter window gives fewer orders to the ads.

Model decides which click gets the credit when the customer clicked several ads:

• Last click: the ad the customer clicked last before buying gets the whole order. This is the most common.
• First click: the first ad gets the whole order. Useful to see which ads bring new customers in.
• Linear: the order is split evenly across all ad clicks. That is why you can see order counts with decimals, for example half an order.

Level decides what a table row is: channel (Google or Meta as a whole), campaign or ad. Google does not send which ad was clicked, so Google is shown per campaign even at ad level.

Below the choices, one sentence explains exactly what you selected.

"How well do we know where the orders came from?"

Every order lands in one of four groups:

The larger the first group, the more you can trust the rest of the page.

Why are so many orders unknown?

The most common causes:

• The customer declined cookies. The pixel then only reports that a purchase happened, without an order number, and Shopify does not store the visit either. The page shows an estimate of how many such purchases there were in the period.
• The order was created manually in Shopify (draft) or in an express checkout such as Vipps. There was no visit in the online store.
• The script was blocked by an ad blocker.
• The order is from before tracking was set up. The page tells you when that applies to your period.

Under the "Unknown" card the group is split into three causes:

• Manual order (draft or point of sale): Shopify says the order was created manually. This is a fact, and such orders get the channel "Manual order" in the list.
• Declined cookies (estimate): the pixel recorded a purchase without consent in the same minute with the same total. This is an estimate, not a certain link.
• No trace: neither. The script may have been blocked, or the order is from before tracking.

Getting more customers to accept cookies is what improves coverage most.

"Which ads got the credit for the sales?"

Each row is a channel, campaign or ad, depending on the level you picked.

A row with ad spend but no orders means Promoosh saw no purchases after clicks on that ad in the period.

The grey rows at the bottom are orders that did not come from a tracked ad click, split by the groups above.

"Did customers buy what the ad showed?"

When a customer clicks an ad for a product, the product is in the link. For each such product the table shows:

• Orders: how many orders where this was the ad that got the credit.
• Bought the same product: how many of them actually contained the product.
• Bought instead: the five most common other products in the same orders.
• Revenue, gross profit, ad spend, ROAS and POAS for those orders.
• Platform's own numbers for the same product.

If a product shows many purchases in Meta but few or none here, the ad attracts customers who buy something else. The product's ROAS in Meta is then misleading, even if the ad may be profitable overall.

The order list

Here you see every order. Search by order number, filter, sort by clicking the column titles, and move columns with the arrows that appear when you hover a column title.

Click an order to see the journey: every ad click and visit in order, how many days it took, the number of sessions, whether the customer is new or returning, the order lines and the amounts with and without VAT. You only see the order number and an anonymous journey, never a name, e-mail or address.

The channels

"(pixel)" after the channel means Promoosh saw the ad click. "(Shopify)" means the channel comes from Shopify's last visit because the pixel saw no click.

Organic search and direct are not the same. Organic means the customer searched and found you. Direct means the customer already knew where you were.

Shopify's visit types

The "First visit" and "Last visit" columns show how Shopify classified the visits. "Google Shopping (paid or free)" means the customer came from Google Shopping, but Shopify cannot tell paid ads from free product listings. "Meta ad (UTM)" means the link was tagged as a Meta ad, but the pixel did not see the click itself. Hover a value for an explanation, or open "What do the channels and visit types mean?" above the list.

Limitations

• The pixel only sees clicks from the same browser. Purchases after someone only saw the ad, or on another device, are invisible.
• Google sends no ad id in the click.
• Unknown means we do not know. Promoosh never fills in a channel that was not measured.