E-commerce and retail
The second order is the whole game.
A first order is expensive and you have already paid for it. A second one costs a message sent at the right moment to somebody who already trusts you. Most shops know this and still do nothing about it, because doing it by hand means somebody watching a list every morning — which lasts about three weeks.
Six funnels, in the order worth building them
Start at the top and stop
when you run out of appetite.
Each one is independent. The first three are worth having within a week; the last three are worth having by the end of the quarter.
1 · Welcome
Trigger: list_join or form_submit. Two messages over ten days — who you are, what they can expect, and one thing worth looking at. Then it stops. This funnel decides whether the next hundred messages get opened, and it is the one most shops skip.
2 · Abandoned basket
Trigger: an api_event your shop posts when a basket goes cold. Wait four hours, send a reminder; wait a day, send one more with a reason to return. Goal: the order. The moment it lands, the rest is cancelled for that person specifically.
3 · After the purchase
Trigger: payment_made. Thank them the same hour. Wait until the parcel should have arrived, ask how it went, and request the review while the memory is warm. Then suggest what people usually buy next — from your own attributes, not a guess.
4 · Replenishment
Trigger: date_field with an offset from the last order — six weeks for coffee, three months for filters, a year for the thing they will not think about until it fails. The single most reliable revenue funnel in retail, and it needs one date.
5 · Win-back
Trigger: segment_enter — three or more orders, none in ninety days. Two messages, then a tag and silence. Sending forever to people who stopped listening damages your sending reputation and tells you nothing.
6 · The VIP letter
Your top hundred customers by value, on a date you choose: a send_letter node, printed and posted. Everybody else in their inbox is a competitor; nobody else is on the doormat. It is the same funnel machinery, with one different node.
Why the goal node matters more here than anywhere.
Retail sequences are short, fast and overlapping. Somebody can be in the abandoned-basket funnel and the win-back funnel in the same afternoon, and the failure mode is not subtle: a discount code arrives twenty minutes after they paid full price.
A goal ends the run for whoever reaches it, and only for them. The basket funnel's goal is the order. The win-back's goal is any order. The review request's goal is the review. Each stops on its own terms, and the person stops hearing from that particular sequence the moment it has served its purpose.
The other half is exit_funnel: entering the post-purchase sequence can eject somebody from the win-back sequence explicitly, because a customer who just bought is by definition not dormant.
- ⚡Trigger
api_event—basket.abandoned, value 214,00 zł. - 2WaitFour hours. Long enough that it is not creepy, short enough to be relevant.
- 3Send e-mailQuiet hours. Held at the message layer until 08:00.
- 4SentThe basket, and nothing else in the message.
- ?Goal checkOrder placed? Not yet.
- 5Wait, then SMSOne segment. Counted before send.
- 6Goal reachedOrder paid. Run exits. The third message is never composed, let alone sent.
Two messages, not fiveStopped by the outcome
The second purchase is cheaper than the first. Considerably.
Getting your shop talking to it
Three connections and you are running.
Contacts in
An export from your platform to start with, imported with a lawful basis you set explicitly. After that, your checkout posts to POST /v1/contacts with your own token, or a form on your site does it for you.
Events in
POST /v1/events for basket.abandoned, order.paid, order.shipped and whatever else your shop knows. Your names, your payload. Each becomes a trigger, a condition and a goal.
Sending out
Your own mailbox or your existing provider. No per-message markup here, and your shop's sending reputation stays attached to your shop rather than pooled with strangers.
If your shop already emits webhooks — most do — pointing them at the events route is usually an afternoon. The attributes you send along with the contact are what make segments sharp later: order count, lifetime value, last category, preferred size. Send them once and every future funnel can use them.
The unglamorous advice: build the welcome funnel first, even though the abandoned basket looks more lucrative. A list that has been warmed by two good messages converts better on every subsequent funnel, and a list that has heard nothing since it signed up will treat your first basket reminder as spam — accurately.
A worked illustration.
A bakery with a small online shop and about four thousand addresses from three years of orders. The list has never been mailed properly, which means it has never been mailed badly either.
The import lands as legitimate interest, because that is what an order history honestly is. MX checking removes several hundred addresses that would have hard-bounced. The welcome funnel goes out to everyone who has ordered in the last year, and asks — politely, once — whether they want the monthly thing, which converts a portion of the list to a real consent basis. Then the replenishment funnel starts running on a date field, and the abandoned-basket funnel starts catching the ones the shop was already losing at nineteen minutes past seven on a Tuesday.
That is three funnels and one import. What it is worth depends entirely on the bakery, and anybody who tells you a percentage is making it up.
Tell us what you send today,
and to how many.
We will show you what it looks like here — usually with the funnel drawn, because that is faster than a paragraph about it.
Open ClickSomo Talk to a person