Mailchimp With Causality Engine: Test the 5-Day Open Window

The short answer: Mailchimp's revenue report and the Email row in GA4 will not match, and neither one is broken. Mailchimp, Intuit's email and SMS marketing platform (entry), credits an order to the last email a contact opened in the previous 5 days or clicked in the previous 30, per its attribution article. GA4 cannot see an open at all. So treat the Mailchimp figure as a claim under that rule and test it with a holdout. Then drop your GA4 Attribution paths export on the start page of Causality Engine, a causal attribution tool that works from that single file (entry). The first finding is free.

Disclosure: I founded Causality Engine and own it, and this page recommends it. Every claim about it links to Causality Engine's own pages.

Three checks before anyone quotes the email revenue figure:

1. Count the windows, not the emails. For example, a list mailed on Monday, Wednesday and Friday gives a contact who opens all three an open window that runs from Monday to the Wednesday after: Friday + 5 days. Ten calendar days, one unbroken claim. Send again the next Monday and it never closes.
2. Start from what is already switched on. Attribution begins by itself once a store is connected, and the windows sit under Analytics, then Analytics Settings (attribution article). GA4 holds the other half.
3. Look for one row. In GA4, open Acquisition, then Traffic acquisition, and find Email in the default channel group. No row means the links were never tagged, and nothing that reads GA4 can see your email.

Mailchimp and the read, row by row

How Mailchimp decides an order was its own

The model is last touch, and the field of play is Mailchimp's own messages. The attribution article says tracking of opens, clicks and deliveries is enabled automatically once you connect an online store, and an order goes to the last message the contact interacted with inside its window. Revenue is shown net of shipping and taxes, so compare it with the store's net sales.

An open is enough. Mailchimp's open tracking article says Apple's Mail Privacy Protection can falsely inflate open metrics. Bot filtering is on by default and hides that activity in reports (bot activity article), but the attribution article does not say whether a filtered open still starts a 5-day window. It does say you decide: under Attribution windows, Email opened and Email clicked can each be switched off or given a different length.

None of this is a flaw. The rule ranks Mailchimp's messages against each other, and it was never built to know about the Google ad the same buyer clicked on Thursday.

Getting Mailchimp clicks into GA4 as Email

Connect a Google account under Integrations, then Google Analytics; it takes Admin access or higher. After that the setting is per message: in Settings & Tracking, check Google Analytics link tracking. In an automation flow it sits under Edit Settings, with Modify UTM Parameters on each email (Google Analytics article).

That article names four parameters your site must accept, utm_source, utm_campaign, utm_medium and utm_term, but prints no default values for a regular email. Mailchimp's classic automation article gives one case: an automation called New User Onboarding arrives as "New User Onboarding / email". Medium is what GA4 needs. Google files a visit under Email when source or medium matches email, and under Unassigned when no rule matches (channel definitions). So send yourself a test, click it, and read the address bar.

Two warnings from the same article. Tracking is not retroactive, so an email sent before the box was ticked stays untagged. And do not stack your own UTM tags on top of Mailchimp's: it recommends choosing one type of tracking.

A worked example of the claim gap

For example, and as an illustration only: a store books 840 orders in a month. Mailchimp's report credits email with 268 of them. Meta's dashboard claims 305 and Google Ads claims 331. GA4's last click gives Email 97.

268 + 305 + 331 = 904 claims on 840 orders, so 904 - 840 = 64 orders are spoken for twice before organic search, referral or direct has claimed a single one. Inside email alone, 268 - 97 = 171 orders sit in Mailchimp's report and somewhere else in GA4. Each one is an open with no click, or a click followed by a later visit through another door.

Nobody lied. Three rules ran over the same orders.

What a causal read adds is a third column, and it comes from the GA4 file, not from Mailchimp. The how-it-works page describes it: each channel's caused share after Direct is split back, next to a last-click rebuild of the same export, a data-health score from 0 to 100, and a next step of raise, cut or hold. If that read puts email well under Mailchimp's 268, do not cut the send plan on its word. Hold a segment out first.

Where this pairing stops

There is no integration between Mailchimp and Causality Engine. The integrations page says the product reads every channel the GA4 export records, with no connector, so email exists for it only as the tagged clicks GA4 kept.

That sets the limits. It reads email as one channel and cannot say which campaign or flow earned its place; Mailchimp's reports and A/B tests do that. It needs a purchase key event and 12 months or more of history (start page). The one-off read page says not to buy when history is thin or broken, and calls a channel under 1% on both methods too small to act on.

Mailchimp's bill follows your contact tier (plans article), and its revenue report is the figure you hold against that bill. With a store connected, opens count from the first send without anyone choosing it, so the default leans toward the larger number. The one person with a reason to narrow it is you.

The free check, then the read

Mailchimp documents the tool for a holdout without calling it one. A percentage split rule randomly sends contacts down path A or B in an automation flow, on Standard or higher. The article uses it to test two emails. Leave path B empty and it becomes a control group, by hand but on a documented rule.

For example, 1,500 contacts enter a cart flow in a month on an 80/20 split: 1,200 get the emails and 300 get nothing. If 138 of the 1,200 order and 30 of the 300 do, that is 11.5% against 10%. The flow caused about 1,200 x 1.5% = 18 orders, whatever the report credits. Fix the end date before you start. For one flow, that answer is enough on its own.

For the whole channel, Causality Engine, which I founded, takes the Attribution paths export on the start page. The preview runs in the browser and is free. The full read is €99 once, refundable within 30 days (pricing, checked on 11 November 2026).

See also

• Mailchimp: plans, limits and complaints that hold
• How to use Mailchimp: the abandoned cart flow, step by step
• Mailchimp on the marketplace: Mailchimp's affiliate program terms
• Causality Engine: what the product reads and costs
• How to use Causality Engine: export, free finding, read, regional test
• Causality Engine on the marketplace: its affiliate program terms
• Omnisend vs Mailchimp: the revenue each one reports
• How to export GA4 data for attribution: the file the read needs

FAQ

Can I connect my Mailchimp account to Causality Engine?

No, and there is nothing to connect. The read takes the GA4 Attribution paths CSV and sees Mailchimp only as the tagged visits inside it, per the integrations page.

Does Mailchimp count an order if the contact only opened the email?

Yes, by default: an order within 5 days of an open goes to that email. The Email opened setting can be disabled or shortened (attribution article).

Why does Mailchimp traffic show as Unassigned in GA4?

Because no channel rule matched the tags. Google's channel definitions need source or medium to match email, so click a test link and read its parameters.