Abandoned cart recovery: from checkout fixes to recovery emails
Why shoppers leave before paying, and how to win them back: checkout fixes, a recovery email sequence and remarketing.
The five automated flows to start with: welcome, abandoned cart, post-purchase, win-back and birthday emails.
The essential email automations are five flows that send themselves based on what a customer does: welcome, abandoned cart, post-purchase, win-back and birthday. Each needs a clear trigger, a timing and a goal. Start with the welcome series and the abandoned cart flow, add the others one at a time, and judge them by the orders they bring rather than by opens alone.
Many businesses send a newsletter when they remember to, and neglect the messages with the most potential: the ones that go out on their own, at exactly the moment the customer is paying attention. Essential email automations do not need a large team. They need a handful of good decisions: what triggers the message, when it goes, what it says and when it stops. Below we walk through the five flows, with structure, timing and content, and then how to build and measure them.
A campaign is a message you send at a moment you choose, usually to a whole list or a segment. An automation is a message, or a series, that sends itself when someone does something: signs up, adds items to a basket, buys, or stops opening anything. A campaign follows your commercial calendar. An automation follows the customer’s moment, which is why automated messages often perform better per recipient.
A few terms used below. Trigger: the event that starts the flow. Segment: a group of people who meet a condition. Exit condition: the rule that stops the flow, for example “has ordered”. Transactional email: a message tied to an order or an account, such as an order confirmation. Deliverability: the likelihood that a message reaches the inbox rather than spam. Double opt-in: confirming an address by clicking a link after signing up.
The table puts them side by side. The timings are indicative and should be adjusted to your buying cycle: a clothing shop moves at a different pace from a B2B service provider.
| Flow | Trigger | Indicative timing | Main goal |
|---|---|---|---|
| Welcome | Joins the list | Immediately, then after 2–3 and 5–7 days | First impression and first action |
| Abandoned cart | Unfinished basket | About an hour, 24 hours, 2–3 days | A completed order |
| Post-purchase | Order placed or delivered | At order, at dispatch, after delivery | Reassurance, a review, a second order |
| Win-back | Inactivity (no click or order) | After a period that fits your cycle | Re-engage or clean the list |
| Birthday | Date of birth or anniversary of the relationship | On the day or shortly before | Relationship and an occasional sale |
A new subscriber is at their most attentive: they have only just signed up and still remember who you are. A welcome series makes use of that moment and typically has three messages:
It helps to ask one preference question on the sign-up form (an area of interest, say), because it lets you send different content from day one. If you use double opt-in, the first message is the confirmation and the series starts after the click. For the right way to collect addresses, read our guide to building an email list in line with the GDPR.
This is the flow with the clearest financial purpose: the customer chose products and left. The usual series has two or three messages within three days, with no discount in the first, and stops when the order is placed. The full structure, the reasons for abandonment and the legal side are in our article on abandoned cart recovery, so here we note only that this flow starts from a real shop event, not from a list.
After ordering, the customer wants to know everything is in order. A good flow mixes transactional and relationship messages: order confirmation, dispatch with a tracking link, usage or assembly instructions, a review request a few days after delivery (never before the product has arrived) and, later, a suggestion of complementary products. The review request is often the most valuable message, because it produces content that helps future sales too.
Keep transactional messages (confirmation, invoice, dispatch) separate from marketing ones, since the consent rules differ. Do not slip an offer into an order confirmation without checking the rules that apply to you.
A big list full of inactive subscribers is not an asset. The more subscribers never open or click, the greater the risk that mailbox providers treat your messages as unwanted, and deliverability can suffer. Gmail, Yahoo and Outlook also expect larger senders (above roughly 5,000 messages a day to their users) to authenticate with SPF, DKIM and DMARC, offer one-click unsubscribe and keep spam complaints below 0.3%; check the current requirements for your own volume. Win-back is therefore also a hygiene measure.
First define “inactive” by your own cycle: for a shop with monthly orders, a few months without activity may be enough, while for rarely bought products the period is longer. Then send two or three messages: one reminding people what they get from you, one with your best content or a small incentive, and one honest “do you still want to hear from us?”. Those who do not respond leave your active lists. This clean-up costs you subscribers on paper, but it improves your statistics and your reputation as a sender.
A birthday message works because it is personal and does not look like a newsletter. It has one condition: you must hold the date, and collecting it must be justified. The GDPR principle of data minimisation says you ask only for what you need. Make the field optional, explain why you want it (“for a small surprise on your birthday”) and store only the day and month if you do not need the year.
If you would rather not collect birth dates, there are alternatives: the anniversary of the first order, of joining a programme or of a subscription. The content stays simple: a short greeting, a clear benefit (a code valid for a limited time, or a gift with the order) and one button. Avoid an over-familiar tone that sounds forced.
Informational note, not legal advice: marketing automation will in many situations require prior consent, a visible unsubscribe link and a clear privacy policy (the GDPR, Regulation (EU) 2016/679, plus the ePrivacy rules; in Romania, Law 506/2004). Check the setup with a lawyer or your DPO before you launch.
The open rate has become a less reliable indicator: some email apps preload the hidden images used to track opens, so the number can be inflated. Treat it as a hint, not as a result. More dependable are click rate, the number of orders attributed to the flow, revenue per recipient, unsubscribe rate and spam complaints.
Compare each flow with your own baseline rather than with figures you have heard elsewhere. To work out the conversion rate from click to order, use the conversion rate calculator. Judge a change after several weeks of data, not after the first two days.
We start with what already exists: the list, the email platform, the shop or CRM data. We then pick the two flows with the clearest impact, usually welcome and abandoned cart or, for a service business, welcome and win-back, get them running, and add the rest only once they work. You will find details of what we do on our email marketing and automation page.
The platform matters less than discipline: a simple automation reviewed monthly beats a complicated flow that nobody looks at any more.
Three is usually a good starting point: one straight away, one after two or three days and one after about a week. The number depends on your sales cycle. If the third message brings unsubscribes and no action, shorten the series and put the emphasis on quality.
Useful criteria are integration with your shop or CRM, a flow builder with exit conditions, segmentation, revenue reporting and GDPR support. Many providers cover the basics, so choose on integration and monthly cost rather than on a long list of features you will never use.
There is no frequency that suits everyone. Automated messages tied to an action are generally better received than those sent for no reason. Watch unsubscribe rate and spam complaints: if they rise after you add a flow, the pace or the content needs adjusting.
It depends on the type of message. Transactional messages tied to an order are treated differently from marketing ones, which in many situations require prior consent. This is a legal matter, so check with a lawyer or DPO and keep proof of consent for each subscriber.
Run a win-back flow of two or three messages, then remove those who do not respond from your active lists. You end up with a healthier base, more accurate statistics and a better chance that your messages reach the inbox.
Event-based flows such as welcome or abandoned cart can show first orders within days, but you need several weeks of data for a solid conclusion. Win-back and birthday flows have lower volume and need more patience.
Laborator
Lab tools that help you put this guide into practice.
Blog
Why shoppers leave before paying, and how to win them back: checkout fixes, a recovery email sequence and remarketing.
Valid consent, double opt-in, lead magnets and list cleaning: a practical guide to a subscriber base you can actually use.
Length, preheaders, spam filters and testing, with good and bad examples and the effect of Apple Mail Privacy Protection on opens.
Tell us where you are now and what you want to achieve — you will get a concrete proposal with services, timeline and a transparent budget.