Triggered email vs campaigns
A campaign is a broadcast: one message, one segment, one send time, judged by what that send produced. A flow is a rule: when a person meets a condition, send this sequence, with these gaps, unless they do something that should stop it. Each recipient enters at their own moment, so the flow sends a trickle every day rather than a burst once a week, and each message arrives close to the event that made it relevant.
That proximity is why flows outperform campaigns per message. A reminder about a basket an hour after it was left speaks to a decision in progress; a newsletter speaks to whoever happens to be reading. The same closeness is the risk: a flow sends on behalf of the business at moments nobody is watching, so a broken template or a wrong trigger reaches people for weeks before anyone notices. Campaigns fail loudly on send day; flows fail quietly.
Both belong in the programme, and the email marketing hub sets out how they share the list, the sending reputation and the calendar. The division of labour is simple: campaigns for news, offers and content that the business initiates; flows for the moments the subscriber initiates, which map onto the stages of the funnel from first contact to repeat purchase. Where the automation extends beyond email into lead scoring, sales handoffs and messaging channels, it is covered in the marketing automation workflows section rather than here.
Flows worth building first
Most businesses can name a dozen flows they would like. The order below is the one that pays back fastest for most, because each flow sits at a moment of high intent, needs data the business already has, and is simple enough to be built well.
| Flow | Trigger | What it does | Build when |
|---|---|---|---|
| Welcome series | New subscriber or account | Confirms the relationship, sets expectations, introduces the offer | First; every list has new entrants |
| Abandoned cart | Items added, checkout not completed | Reminds, resolves objections, brings the shopper back | As soon as the shop can identify the shopper |
| Browse abandonment | Product viewed repeatedly, no cart | Lighter nudge towards the product or category | After abandoned cart is working |
| Post-purchase | Order completed | Thanks, sets delivery expectations, asks for a review later | With the first repeat-purchase goal |
| Replenishment or renewal | Time since purchase for consumables or subscriptions | Reminds before the product runs out or the term ends | Where the product has a natural cycle |
| Re-engagement | No clicks or visits for a defined period | Asks whether to stay; removes those who do not respond | Before list hygiene becomes a deliverability problem |
| Win-back | Lapsed customer past the usual repeat interval | A reason to return, with any incentive kept for the last message | Once customer lifetime value (LTVTotal value a customer brings over time.) by cohort is known |
Build one flow at a time, run it until its reports are trusted, then build the next. A programme with two flows that work beats one with seven that nobody has checked.
Welcome, abandoned cart, metrics: the core guides
The core pages cover the practical work of this section.
- Building a welcome series is a step-by-step tutorial from the goal and entry rules through the first email, spacing, content per message, personalisation by sign-up source, exits, tests and launch. It is the flow to build first and the pattern every later flow reuses.
- Building abandoned cart flows is the e-commerce tutorial: identifying the shopper, cart and browse triggers, timing, what each email contains, an incentive policy that does not train people to abandon, exits on purchase, and frequency rules. The cart abandonment rateShare of shopping carts created that did not become an order. is the measure that shows how much intent those flows have to work with.
- Email metrics and testing is the measurement guide for both: metrics by layer, what opens can and cannot tell you now, clicks and revenue per emailEmail-attributed revenue divided by the number of emails delivered., holdouts for flows, and how to run an A/B TestControlled comparison of two variants. with enough volume to read.
The tutorials assume a working list with a recorded consent basis, which the fundamentals section covers, and they are platform-agnostic: every mainstream sending platform offers triggers, delays, conditions and exits under its own names, and the tutorials describe the logic rather than any one interface.
Further guides
Further tutorials build flows for particular moments: after an order, between a software sign-up and its first real use, and when a product someone wanted returns or falls in price:
- Building a post-purchase email flow — a step-by-step tutorial for the flow after an order: working beside order emails, help with what was bought, the review request and the next purchase.
- Building onboarding emails for a software product — a step-by-step tutorial for the activation goal, product-event triggers, one job per email, the trial end, exits and testing.
- Building back-in-stock and price-drop alerts — a step-by-step tutorial for alerts a customer asked for: the request and its wording, stock and price rules, who is told first, expiry and measurement.
Triggers, timing and exits
Every flow is three decisions. The trigger is the event that puts a person in: a sign-up, a cart event, an order, a date, a period of silence. It needs to be specific enough to mean one thing, because a trigger that fires on several kinds of event produces a message that fits none. The timing is the gap between the trigger and each message, and the gaps between messages: close to the event where intent is fresh, wider as the sequence goes on. The exit is the condition that takes a person out: they bought, they replied, they unsubscribed, they entered a flow that takes precedence, or the sequence ended.
Exits are where most flows are weakest. A welcome seriesThe automated sequence of emails sent to a new subscriber in the days after sign-up. that keeps selling to someone who bought on day one, or a cart reminder that arrives after the order confirmation, is a flow with a missing exit. Before any flow goes live, list every event that should stop it and check that the platform evaluates those events before each send, not only at entry. Priority between flows matters as well: when a person qualifies for two flows at once, decide which one wins and whether the other is delayed or dropped.
A homeware shop fixes its exits
A homeware shop ran a three-email welcome series and a three-email abandoned cart flow, each with exits only at entry. A review of 6,200 flow sends in one month found that 740 welcome emails had gone to people who had already placed an order, and 310 cart reminders had been sent after the cart in question had been purchased, including 90 that arrived within an hour of the order confirmation. Adding a purchase exit evaluated before each send, and a rule that the cart flow pauses the welcome series for anyone in both, removed those sends entirely. Monthly flow volume fell by about 17%, support emails asking why the shop was reminding people about something they had already bought stopped, and the shop's conversion rate from the remaining cart reminders rose because the denominator no longer included people who had already converted.
Governance for unattended flows
A flow runs without a person pressing send, so the controls that a campaign gets at send time have to be built in. Four practices cover most of it. Ownership: every flow has a named owner who reviews its reports and is told when a template or product it references changes. Review cadence: a calendar entry, monthly for flows with volume and quarterly for the rest, to read the figures, open each email on a phone, and click every link. Change control: edits to a live flow go through a test send and a dated note of what changed, because the person diagnosing a drop in six months will need to know. And a kill switch: a documented way to pause every flow at once when the shop's catalogue, pricing or legal position changes.
Measurement belongs to governance too. Each flow has a goal metric agreed at build time, usually conversion rate (CVRShare of visits or clicks that convert.) to the action the flow exists for, alongside the unsubscribe and complaint figures that show whether the flow is wearing out its welcome. Those key performance indicators (KPIA metric tied directly to an objective.) are read against a holdout where volume allows, which the metrics guide explains, because a flow that reaches people about to buy anyway will look excellent and add little. Governance for automation that spans channels and systems, including approval, naming and audit trails, lives in the workflows section of the automation root.
Common automation problems
Most flow problems show up as one of a few symptoms. The table maps each to the page and section that deals with it.
| Symptom | Usual cause | Where to look |
|---|---|---|
| Welcome emails arrive days after sign-up | Trigger tied to a batch sync rather than the sign-up event | Welcome series, trigger and entry rules step |
| Cart reminders sent after purchase | Exit evaluated at entry only | Abandoned cart flows, exits on purchase step |
| Regular customers expect a discount | Incentive in every cart email | Abandoned cart flows, incentive policy step |
| Flow looks profitable but revenue does not move | No holdout; flow reaching people who would buy anyway | Email metrics and testing, holdouts section |
| Rising unsubscribes from one flow | Too many messages, too close together, or stale content | Welcome series, sequence and spacing step; diagnosis: unsubscribe rate spiking |
| Subject line test called a winner too early | Insufficient volume for the difference seen | Email metrics and testing, A/B testing section |
| Same person in two flows at once | No priority rule between flows | This page, triggers, timing and exits; diagnosis: subscribers receiving duplicate emails |
| An unsubscribed person still receives a flow's emails | Suppression checked only at entry, or an opt-out that never reached the platform | Unsubscribed contacts still receiving email |
| Post-purchase emails arrive before the order has shipped, or repeat the order confirmation | Messages timed in fixed days rather than from the order's progress; the flow never mapped against the order emails | Post-purchase flow |
| Trial users receive the same onboarding emails whatever they have done | Emails sent on fixed days rather than triggered by the step a user has not taken | Software onboarding emails |
| More shoppers are told an item is back than there is stock to sell | No stock rule; every request notified at once | Back-in-stock and price-drop alerts |
