Lifecycle and automation flows

What this actually solves

The most expensive thing in most online businesses is not the traffic that never converts. It is the traffic that nearly converted. Someone who added to a cart and left, someone who signed up and was never spoken to again, someone who bought once and has not been given a reason to come back. That demand has already been paid for. It is sitting there, close, and nothing is reaching for it.

Owners feel this one directly, which is why it sells more easily than almost anything else in content. They can watch the cart abandonment figure in their own analytics. They know roughly how many people put something in a basket last month and how many actually paid. The gap is visible and it is money.

Flows are the answer to it, and they are the most engineering-shaped work in this category. There is copy in them, but the part that decides whether they work is the logic — what fires them, how long the gaps are, when they stop, and crucially who is kept out of them. Written once, they run for a year. Written carelessly, they send an abandoned cart reminder to someone who has already paid, and you hear about it from your own customer.

What we do, and in what order

Nothing gets written until the map is agreed. Copy is the cheap part to change; logic is not.

1. Read the data and find the drop-offs. Where customers enter, what path they take, where they fall out, and what the existing setup already sends. This produces the arithmetic we show you — built from your sessions, cart rate, conversion rate and average order value, not from a benchmark someone published. Two to four days.
2. Agree the flow list and the order of build. Usually welcome, abandoned cart and post-purchase first, with the rest ranked by what they are plausibly worth against how long they take. You see the reasoning, including the ones we recommend not building yet. One to two days.
3. Map every flow before writing a word. For each — the trigger, the number of emails, the delay between each one, the purpose of each in a single line, the exit condition that stops the sequence, and the exclusions that keep the wrong people out. This document is signed off by you and it is the thing that prevents the most expensive failures. Three to five days.
4. Write the copy and build it in the ESP. Subject lines and preheaders per email, body copy written to stay true months from now, then the triggers, conditions, delays and suppression rules built and connected to the store data. One to two weeks, depending on how many flows.
5. Walk every flow live, end to end. A real address, a real cart genuinely abandoned, a real test order placed and refunded. Every email received, every link clicked, every exclusion verified by trying to break it. This takes under an hour per flow and it is the step that separates a build that works from a screenshot of one. Half a day to a day.
6. Review at day thirty. Open, click and revenue per email, where in each sequence people leave, which exclusions proved too broad or too narrow, and the fixes applied. Half a day, thirty days after launch.

Tools — Klaviyo where there is a Shopify or similar store, ActiveCampaign or HubSpot for B2B, MailerLite for simpler builds, and a plain document or Whimsical for the flow maps, because the map needs to be readable by you rather than pretty.

What moves the price

This is a fixed build rather than a retainer — $600–$5,000, 2–5 weeks from the data review to the live walk-through, with the day-30 review after it. Three things decide where a project lands in that range.

How many flows and how many emails inside each. Three core sequences with three or four emails apiece sits near the bottom of the range. A full lifecycle system with browse abandonment, win-back, replenishment, review requests and segmented variants sits at the top, and it is genuinely several times the work rather than a few extra sends.

The platform. Klaviyo can do more and therefore takes longer to configure properly — product feeds, conditional splits, purchase-based segments. The same six flows in MailerLite are quicker to build and less capable when they run.

Whether design and extra channels are in scope. Copy-only flows in the ESP's default template are fast. Designed templates with product blocks, or an SMS and WhatsApp layer with its own consent capture and quiet hours, are each a separate body of work.

What usually goes wrong

The flow is built and never walked. Everything is configured, a screenshot goes to you, the invoice goes out. Two weeks later it emerges the abandoned cart sequence is reminding people to complete a purchase they already completed, because the exclusion for recent purchasers was never added. The customer complains to you, and you associate the embarrassment with whoever built it. Walking it takes under an hour per flow. There is no reason ever to skip it.

Nobody defines when the sequence stops. Exit conditions are the part beginners forget entirely. Without them, a welcome series keeps arriving after someone has bought, a win-back keeps chasing a customer who came back last Tuesday, and a person who qualifies for two flows at once receives three emails before lunch. The fix is boring and it is written into the map before the build starts.

Campaign copy gets written inside a flow. A flow is evergreen by definition. Put "this week only", a seasonal reference or a specific discount deadline inside one and it is wrong within a month and still sending in March. Every line in an automated sequence has to be true a year from now, and offers inside flows are either genuinely permanent or pulled from a field that can be updated in one place.