The seven email automation flows, ranked by revenue

Omnisend measured 20 billion campaign emails and 470 million automated sends across more than 27,000 brands last year. The campaigns, meaning the newsletters and promotions somebody schedules by hand, returned $0.155 per email sent. The automated sends returned $3.41 per email, a 22-fold difference on the same lists selling the same products.

The difference is timing. A campaign goes out when your calendar says so, and reaches most people while they are thinking about something else. A flow goes out when the customer does something: joins your list, abandons a checkout, buys, goes quiet. The seven below are ranked in build order, each with its benchmark, its timing, its message, and its service-business version. A plumber has no cart, and most writing on this subject pretends otherwise.

Rank flows by revenue per recipient, not open rate. Apple's Mail Privacy Protection loads images for the reader whether or not they open the message, so a share of every open rate you have seen since 2021 belongs to a machine. Brevo's 2026 benchmark is one of the few that publishes the average open rate both ways, with those automatic loads counted and with them stripped out. Revenue per recipient carries no such distortion.

One caution about that headline number. Omnisend publishes two 2026 pages that disagree: the benchmark report says $3.41 against $0.155, its statistics page says $2.87 against $0.18. Read the gap as 16 to 22 times, and every figure here as an average across 27,000 unlike businesses rather than a forecast for yours.

None of it works without the setup underneath. An authenticated sending domain and a list of people who opted in are the price of entry, covered in our guide to setting up email marketing. The case for building the list sits in why owned audiences beat rented reach.

The welcome flow, and it is not close. Omnisend's table below ranks automation types by revenue per email sent. Back-in-stock alerts top it at $9.14, but only a business that carries inventory and runs out of it can use that one. Welcome comes next at $6.16, and any business with a form can build it.

Two things stand out. The transactional messages, order and shipping confirmations, earn well because people open them at rates almost no marketing email reaches, which makes them the most wasted real estate in email. And customer reactivation, the flow agencies sell hardest, sits near the bottom of every column that matters.

So the build order runs like this.

1. Welcome, week one. The highest revenue per email of any flow every business can build.
2. Abandoned checkout or quote, week one. Someone told you what they wanted and then stopped.
3. Post-purchase or post-job, week one. Lower per email, and it runs on people who already paid you.
4. Review request, week two. Fourth by revenue, second by real value if you sell services locally.
5. Lead nurture, next quarter. The most expensive to build, and the slowest to pay back.
6. Cross-sell or replenishment, next quarter. Modest per email, close to free to run, with a ceiling set by your catalog.
7. Win-back, last. The weakest automation type measured, and more useful as list hygiene than as revenue.

A business that builds only the first, the second and the fourth captures most of the money on the table.

Welcome emails earn $6.16 per send at a 35.53% open rate and 2.11% conversion. Omnisend's separate welcome-series guide puts welcome conversion at 3.7%, above its own benchmark table, so read the honest range as two to four percent. The flow works because of when it arrives, seconds after a stranger voluntarily hands over an address.

Send three emails over about seven days. No published benchmark settles the number, so judge it by the work. One email has to introduce you, prove you are worth trusting, and ask for the sale, and it does all three badly. Split across three, each message carries one job.

The three emails, and what each one says

1. Within minutes. Deliver whatever you promised at signup, the code or the guide, in the first line, with one link to what happens next.
2. Day two or three. A store tells its founding story and shows what people buy most, and a service business answers the objections that kill deals.
3. Day five to seven. Make the ask. A store pushes product against the incentive from email one. A service business puts the booking link at the top and the bottom.

For a contractor, a law firm or a dental practice, the same flow runs after a quote request. The buyer is deciding whether to let you into their house or their finances. Email one confirms who is calling and when, email two answers the objections, and email three makes booking one click. Businesses in construction and the trades win or lose these jobs in the hour after the form arrives, and the flow holds the line until a human calls.

Baymard Institute pools 50 separate studies and puts documented cart abandonment at 70.22%. Seven of every ten people who add to a cart leave without buying, and the reasons they give are about the checkout, not the price. Extra costs at the end, shipping and tax and fees, account for 40%. Being made to create an account accounts for 18%, and a slow checkout another 17%.

Read those numbers before you write the emails, because no email fixes a shipping surprise. If 40% of your abandoners left over a cost they first saw at step four, the repair is showing that cost at step one. The flow only catches what the checkout leaks, which makes the checkout the first thing to audit on an e-commerce store.

The flow runs at one hour, 24 hours and 72 hours. The widely quoted 20.3% conversion figure for the first hour traces to a vendor study no longer reachable at source, so treat the cadence as practitioner consensus. Baymard reports that 42% of US online shoppers have abandoned a cart because they were browsing, and a same-day discount to a browser buys back a sale a reminder would have won. Save the incentive for email three, on carts above your average order value.

The service version: the abandoned quote

A plumber has an estimate sent Tuesday that nobody answered by Friday, the same event in different clothes. The trigger is a quote sent with no response, and the cadence stretches because the decision is bigger: 48 hours, five days, twelve days. Email one repeats the price and the scope in plain words. Email two answers the objection the price raised, usually warranty, timeline, or who does the work. Email three offers a smaller version of the job or a call, then stops, and it all runs on quotes you already paid to produce.

Order follow-up earns $1.75 per email at a 47.70% open rate. Lower than welcome, and it compounds, because it runs on people who already paid you. A store spends days one through seven on shipping and setup content, then returns at the product's reorder interval. A service business uses the same window to say what happens next, how to care for the work, and who to call at nine at night.

The review request earns $1.14 per email at a 49.17% open rate, which understates it badly. BrightLocal surveyed 1,002 US consumers in 2026 and found 78% had been asked for a review in the past year, and 83% of those asked left one. The return lands in map rankings and close rates, not in the email's own conversion column.

Reviews also expire in practice, because 74% of consumers seek out reviews written in the last three months. A business sitting on forty five-star reviews from 2023 has, for buying purposes, none. That is the argument for automating the ask instead of running a review push twice a year. Send it three to seven days after the job, one ask with one link.

Lead nurture has no clean public benchmark, and behaves like a slower welcome flow. It suits purchases people think about for weeks: professional services, business software, a roof, a kitchen. Mail every week or two, teach rather than pitch, and keep one low-friction call to action in every message. It ranks fifth because it costs the most to build and pays back the slowest, though a single five-figure job covers the build.

Cross-sell and replenishment earn $0.95 per email at 0.87% conversion. Modest, close to free to run, and the only flow whose ceiling is your catalog rather than your traffic. The service version is the annual plan, the maintenance contract, the second property.

The flow that is oversold

Customer reactivation is the weakest automation type Omnisend measures. It opens at 33.11%, clicks at 1.99%, converts at 0.54%, and earns $0.51 per email, roughly a twelfth of what welcome earns. It also unsubscribes 0.68% of the people it reaches, half again as many as the abandoned-cart flow that earns seven times more per send. The damage lands where you cannot see it: you mail people who ignored you for a year, and inbox providers learn to filter everything else you send.

The 10 to 15% reactivation rates quoted across deliverability blogs cite other blogs rather than a study. Use the measured number, and build the flow for the thing it is good at, which is sorting. Send three emails, then suppress everyone who ignored all three: list hygiene with an occasional sale attached.

The trigger matters more than the tool. An estimate sent 48 hours ago and never accepted is an event that lives in Jobber or ServiceTitan, not in Mailchimp, and owners who try to rebuild it inside an email platform abandon the project. Automate from the system that owns the event.

Two rules cut through that list. Selling products through Shopify, start with the built-in flows or Omnisend, and move to Klaviyo when segmentation costs you money. Selling services, pick the tool that already holds your jobs and quotes, and add an email platform only when its follow-up runs out of room. Connecting the two is ordinary automation work, and it stalls these projects more often than the copy does. We do that wiring inside our email marketing engagements, so tell us what broke.

Every number here is an average, and an average across 27,000 businesses describes none of them. Klaviyo is the only vendor publishing benchmarks in bands. A store under $1M in revenue with an average order value between $44 and $83 sees abandoned-cart revenue per recipient of $0.86 at the 25th percentile and $2.91 at the 75th. Most benchmark posts quote $3.65 for the same flow, an all-brands mean that includes stores with far larger order values than yours.

Owners read the $3.65, measure a third of it, and rip out a flow that was performing normally. Find the band that matches your revenue and order value first.

The $36 myth

The claim that email returns $36 for every $1 spent is the most repeated line in this trade. Litmus, usually credited with it, published something else: a survey of about 500 marketers guessing at their own results. 35% get back $10 to $36, 30% get $36 to $50, and 5% get more than $50. The page now carries a banner saying it is more than two years old, and a third of those marketers sit below the number everyone quotes.

Omnisend's 470 million automated sends did not come from 27,000 marketing departments working harder. They came from sequences somebody set up once and left running. The seven above take two weeks to build, and after that they go out at one in the morning, while the competitor's newsletter sits in a draft folder waiting for Tuesday.