Flodesk With Causality Engine: Check What the Clicks Sold

The short answer: Flodesk shows who clicked. Shopify shows who paid. Nothing on either screen joins the two, so the sales figure for email is one you end up building yourself. Flodesk, the design-led email tool for creators and small brands, documents opens and clicks and stops there: its abandoned cart workflow treats a Shopify checkout as abandoned after 10 minutes and waits 1 hour by default before it writes (abandoned cart guide), and no help article counts the orders that follow. Tag the links by hand so GA4 files them as Email. Then Causality Engine, a causal attribution tool that reads GA4's export, can weigh email against the other channels, starting with a free first finding at app.causalityengine.ai/start.

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

Before clicks get multiplied into sales:

1. Do the tempting sum, and label it a guess. For example, 300 clicks at an assumed 4% store conversion rate is 300 × 0.04 = 12 orders. The 4% belongs to the whole site. Email readers are not the whole site.
2. Use the builder Flodesk gives away. Its free Campaign UTM builder turns a product URL into a tagged one, with source, medium and campaign as required fields.
3. Name the row you will watch. After the next send, the Email row in GA4's Traffic acquisition report is the count to set beside your 12.

Flodesk and the read: every fact with its page

Tagging comes first, because Flodesk leaves it to you

Flodesk's sending guide explains that each link is rewritten to pass through its tracking system, which records the click and then forwards the reader to the destination you typed. That feeds Flodesk's own analytics. It tells GA4 nothing. The help centre articles read on 11 November 2026 describe no setting that appends UTM parameters, and the integrations page counts 16 direct integrations with no analytics tool among them.

So the tag is typed by you, once per link. Google files a visit under Email when its source or medium is email, e-mail, e_mail or e mail (default channel group). The builder's own FAQ offers 'newsletter' as a source and 'email' as a medium, and that pair passes. A medium of 'newsletter' does not.

Three traps, each from Flodesk's pages:

• Capital letters. The builder's FAQ says 'Newsletter' and 'newsletter' are tracked as two sources, and offers a casing option in Pro mode.
• Renaming later. Data collected under old parameters does not move to new ones, so fix the naming before the first send.
• Test emails. Links in a test send expire after a few hours (sending guide). Check tags on a live email.

One gap: the abandoned cart guide does not say whether the cart button can carry tags. Send yourself an abandoned cart email and read the address bar.

What Flodesk counts, and where the count ends

A sent email reports Sent, Delivered, Opens and Clicks, the last defined as how many subscribers clicked a link (email results). Flodesk is candid about the last two. Its Mail Privacy Protection article says open rates are no longer a reliable measure, and calls click data "a directional indicator, not a perfect measure" because security scanners follow links too.

The subject line A/B test runs 24 hours and picks its winner on open rate alone (A/B testing). The Shopify connection syncs customers and starts purchase follow-up and abandoned cart workflows (Shopify guide); under the abandoned cart guide, a buyer is removed from the workflow once the purchase is detected. Orders and revenue appear on one analytics page only, Checkouts, which covers Flodesk's own checkout pages (Analytics overview). For a connected Shopify store, the documentation ends at the click.

Flodesk says as much. The privacy article names purchases as the strongest signal and adds: "Adding UTM parameters can help you measure this in analytics tools."

A worked example: 300 clicks, 95 orders

For example, with made-up numbers chosen to show the sums, not taken from any store: a candle shop's campaign is delivered to 6,000 subscribers and 300 click, a click rate of 300 ÷ 6,000 = 5%. Over the next three days the shop books 95 orders.

Add the three claims: 12 + 52 + 44 = 108, which is 108 - 95 = 13 more orders than the shop booked. The hand sum also overshoots GA4's Email row by 12 - 9 = 3. And 31 orders sit under Direct, a row that names no sender at all.

That Direct row is where a causal read starts. Per its how-it-works page, it splits Direct back to the channels that sent those buyers, then shows each channel's caused share beside a last-click rebuild, with a data-health score from 0 to 100 and a next step. If it rates Email above your 12, do not send twice as often on that alone. Test on a held-out group first.

What the pairing cannot do

• No purchase key event in GA4, no read. The start page asks you to pick that event and to export 12 months or more.
• A list that sends little traffic can land under the floor: below 1% of credit on both methods, a channel is called too small to act on, and traffic with no tracked source is flagged as Unassigned, with the advice to fix tagging first (pricing page for agents).
• The read does not design, send or segment. Which subscriber clicked which button stays in Flodesk.

Flodesk is paid its plan fee whether an email sells or not, and the dashboard that comes with it ends at the click: a number that needs no store connection to produce and never has to be reconciled with an order. Reconciling is left to whoever pays the bill.

The free test: hold a tenth back

Flodesk documents the exclusion. A workflow's Settings carry an Exclude subscribers option that takes segments (workflow exclusions), and its unengaged subscribers guide tells you to exclude a segment from campaigns as well. The random pick is by hand: export the list, shuffle it in a spreadsheet, and add every tenth address to a segment called Holdout. Exclude it from four promotional sends. Then match orders to both groups by customer email.

For example, 5,400 mailed subscribers give 81 buyers and 600 held-out subscribers give 6: 81 ÷ 5,400 = 1.5% against 6 ÷ 600 = 1.0%. The emails added (1.5% - 1.0%) × 5,400 = 27 orders.

That answer is free and stands alone. When the question widens to email against ads and search, export the Attribution paths CSV and drop it on the start page, where the preview is worked out in your browser at no charge. Causality Engine, which I founded, sells the full read of the same file at €99, once, excluding VAT, refundable for 30 days (pricing, checked on 11 November 2026).

See also

• Flodesk: the tool, its plans and its limits
• How to use Flodesk: delivering a lead magnet with a workflow
• Flodesk on the marketplace: Flodesk's affiliate program terms
• Causality Engine: what the read is and costs
• How to use Causality Engine: from export to budget decision
• Causality Engine on the marketplace: the affiliate program terms
• Flodesk vs Klaviyo: design-led email against ecommerce email
• How to find what Direct traffic is hiding: the Direct row, explained

FAQ

Can Flodesk send its data to Causality Engine?

No, and it does not need to. The two products never touch. Causality Engine's integrations page describes a product that reads the channels in a GA4 export, so Flodesk's emails count once their links are tagged and GA4 has recorded the visits.

Does Flodesk report revenue from Shopify abandoned cart emails?

Not in its help centre. The abandoned cart guide covers the trigger, the delay and removal after purchase, and the Analytics overview lists orders and revenue under Checkouts only.

Flodesk already tracks clicks. Why add UTM tags?

Because the click is recorded on Flodesk's redirect, before the store page loads. GA4 sees only the arriving visit, and Google's (direct) / (none) page says links without UTM parameters lose their source.