„Our Shopify analytics already shows marketing attribution.“
For orders placed online, and it does it well. This is about the orders placed in a shop or over the phone, which those reports cannot see at all.
Für Shopify
Export your orders, upload your calls and leads beside them, and see which marketing produced the sales that never had a click on them.
Kein API-Schlüssel Nichts in Shopify zu installierenNichts zu installieren Keine Karte nötigOhne Karte
CloseRev reads a Shopify export of orders — the customer's email or phone, the order total and the date — and matches the ones with no click behind them against your call and lead records. Online checkout is already attributed and needs nothing from us; the point of this page is Shopify POS, phone orders and draft orders, where somebody researched online and bought where no pixel could follow. Orders that match no lead are reported as Direct / Unknown rather than credited to the last campaign that ran.
Zuletzt am 24. September 2026 anhand der eigenen Dokumentation von Shopify geprüft.
Die Lücke
Your online orders are measured to the penny. The customer who compared three sofas on your website and then rang to order is invisible to every report you have.
Was verkauft wurde, an wen und für wie viel.
Der Klick, das Keyword, der Anruf und was jedes davon gekostet hat.
Die Datei
Drei Dinge tragen die Zuordnung: wer, wie viel und wann. Alles andere ist optional und ändert nur, wie sich der Bericht aufschlüsseln lässt.
On the order's customer. Online orders always carry an email; point-of-sale and phone orders carry whichever your staff captured, which is why both columns belong in the file.
What the order took, after discounts and including shipping if you want it counted. Consistency matters more than which you choose.
When the order was placed. For draft orders sent as a quote, the date it was paid is closer to the decision.
Online store, point of sale, draft order, and which shop. This is the column that separates what already has attribution from what does not, and it is the most important optional field on this page.
Schritt für Schritt
Geschrieben für alle, die Shopify im Nachbar-Tab geöffnet haben. Berichtsnamen unterscheiden sich je nach Edition, deshalb sagt jeder Schritt, wonach Sie suchen.
One row per order with a customer contact detail, the total, the date and — if at all possible — the channel or location it came through. Without the channel you cannot tell the already-attributed orders from the ones this exists for.
Online orders with a click behind them are already measured properly by your ad platforms. The value here is in point-of-sale, phone and draft orders, so isolate those first and read them on their own.
Call tracking, the contact form, showroom enquiry logs, quote requests. Contact detail plus the source, campaign or keyword.
Considered purchases are researched for weeks before somebody walks in or rings. A window as short as the ad platform's will miss exactly the orders this is for.
The join is the email and the phone, normalised on both sides, with weak links flagged rather than counted.
Was zurückkommt
Umsatz nach Kanal, die Zahl der Verkäufe hinter jedem Wert und eine ehrliche Kategorie für alles, was niemand zuordnen konnte. Beispielzahlen aus dem durchgerechneten Beispiel der Seite E-Commerce & Einzelhandel – nicht aus einem Shopify-Konto.
| Kanal | Anteil | Verkäufe | Umsatz |
|---|---|---|---|
| Google Ads | 84 | $98,000 | |
| Meta Ads | 61 | $53,200 | |
| E‑Mail-Marketing | 47 | $44,800 | |
| Direkt / Unbekannt | 72 | $84,000 |
Nicht zugeordnete Verkäufe bleiben unter Direkt / Unbekannt. Sie werden nie auf die bezahlten Kanäle verteilt, damit die Summe besser aussieht.
Das Argument
A straightforward online purchase is one of the best-measured events in commerce. The click, the session and the checkout are all in one place, the ad platform sees the conversion, and nothing here improves on that.
What is not measured is everything else Shopify now runs: a point-of-sale order in a shop, a phone order taken by a person, a draft order sent as a quote and paid days later. The customer researched online — that is where they found you — and then bought somewhere the pixel does not reach.
For merchants with a physical presence or a considered product, that is not a rounding error. It is frequently the larger half of the revenue and the half that decides whether the advertising pays.
Under fifty, people buy on the spot. Over a few thousand — furniture, equipment, jewellery, bikes, appliances — they research online, then come in to see it, then buy.
The result is a systematic bias in every report the merchant reads: the cheap impulse categories look efficient because their whole journey is online and measurable, and the expensive considered categories look like they do not work.
Budget follows the measurement rather than the money, and it does so quietly for years. Matching offline orders back to the enquiries and calls that preceded them is the correction, and it usually moves spend toward the categories that were carrying the business all along.
An online order cannot exist without an email. A shop order can, and often does — the transaction completes with a card and nothing else unless somebody asks.
That gradient decides your matched share. Merchants who capture an email at the till for the receipt match well; merchants who do not will see a large Direct / Unknown and should read it as a property of their till process rather than as a finding about their marketing.
It is also the cheapest thing on this page to fix. Asking for an email at the counter changes what can be measured more than any change to the advertising would.
Retail lives on repeat purchase, and a customer produced by a campaign two years ago may have bought fifteen times since. Crediting all of it to that campaign makes it look extraordinary; crediting none of it understates a channel producing genuinely loyal customers.
Neither answer is right for every merchant, so the report does not choose. First orders and subsequent orders are reported separately, and the decision about which to spend against stays visible and yours.
Berechtigte Fragen
For orders placed online, and it does it well. This is about the orders placed in a shop or over the phone, which those reports cannot see at all.
Then you probably do not need this. If you take any phone orders or send draft orders as quotes, those are the ones worth reading.
Then the matched share will be low and the report will say so. Capturing an email at the counter is usually the highest-return change a shop can make here.
No. There is no app to install and no scope to grant. It reads a file you exported, which is also why it cannot touch your storefront.
Orders with a customer contact detail, the total, the date and the sales channel or location where possible.
You can, but they are already attributed properly by your ad platforms. The value here is point-of-sale, phone and draft orders.
On the sales channel or location column, which is why it is the most important optional field on this page.
Yes, and it is the main reason to do this. A POS order with a customer email or phone matches like any other row.
They match too, and the paid date is usually closer to the decision than the created date.
Include them and they reduce a channel's revenue, which is the honest figure.
They are reported separately from first orders, so you decide whether two years of repeat purchase should credit the advert that started it.
Direct / Unknown. They are never matched on a name, because names are not unique.
Yes, where the export carries one. A multi-site retailer's blended figure is usually true of none of its shops.
A contact detail, an amount and a date, plus a channel if you include one. No line items, no addresses, no payment details, no browsing history. Encrypted in transit and at rest and deleted with the import.
Revenue by channel for the orders that had no click of their own, the share matched, and everything unmatched kept visible.
Nach Branche
Wie der Bericht aussieht, sobald der Export drin ist – für jede Branche eigens geschrieben.
Andere Systeme
Mehrere Systeme im Einsatz oder im Vergleich? Die Methode ist dieselbe, die Spalten sind es nicht.
Shopify und die anderen Produktnamen und Logos auf dieser Seite gehören ihren Inhabern und dienen nur dazu, die Software zu kennzeichnen, aus der eine Datei stammt. CloseRev ist mit ihnen nicht verbunden oder von ihnen empfohlen und verbindet sich mit keinem davon: CloseRev liest eine Datei, die Sie exportieren.
Heute starten
Nichts in Shopify zu installieren, kein API-Schlüssel, und Sie müssen bisher nichts getrackt haben. Das Vorjahr funktioniert so gut wie dieser Monat.