Ecommerce email marketing starts with five boring flows

Two of the three store owners we found who run their own ecommerce email marketing (the automated flows a shopper's actions set off, plus the campaigns a store schedules to its own list) report a smaller slice of revenue from it than agencies say it should bring. We gathered the shares people stated about email marketing for ecommerce and sorted them by who was doing the counting.

Two of the three owners put email at 10% and 15%. The third, a dropshipping store owner, reported 35% and called those sales "basically free" next to paid ads. Set aside the before picture, and the agency and auditor figures run from 26% to 53%. The one low agency number, roughly 8% for an electronics brand doing about $500,000 a month, is a before picture. The agency used it to show what "we already do email" looks like.

The 20-30% band that gets quoted as the target comes from the same side of the table. An email auditor who says they've audited 100-plus brands gives "target is 20-30%. Sometimes 35%+" as the first check on an account. Another agency says the average brand it audits sits at 15-20% for email and SMS combined, "when it should be doing 30 to 50%". Both sell email work.

That doesn't make them wrong. It does mean the targets quoted to stores in our research came from people who get paid to raise them. And every figure in the table is a share someone reported. None of them was audited. Further down, even email specialists disagree about what that number proves.

A retention agency described the email account it inherited from a cannabis delivery brand, and it reads like a list of everything missing. The brand only emailed past buyers. Its welcome flow was three emails, mostly about the discount. No browse abandon, no cart abandon, no checkout abandon, no proper post-purchase flow. Campaigns went out two or three times a month, and the signup popup was a spin-to-win that waited 30 seconds before showing up.

The agency started with the form. By its count, the popup's conversion rate went from 6.6% to 9.5%. The new form also collected shopping intent, so the welcome emails could be relevant to what a person came to buy. Then it built the five missing flows: welcome, browse, cart, checkout and post-purchase. Then it warmed up the sending domain and slowly send campaigns more often, filled with product education, founder-led pieces and trust content, "not just discount after discount".

Thirty days later the agency reported $21,883 in attributed email revenue, with flow revenue up 265% and campaign revenue up 105%. Those are the agency's figures for its own client, as the email platform counted them. Its summary was plainer than the numbers: "Most accounts I see aren't even doing the boring stuff properly yet."

What goes inside the emails can be plainer than you'd expect. An owner who has run several online-store brands for 12 years, some B2B and some D2C, said they'd A/B tested marketing emails across them the whole time: "The ugly, short, 1-link plain text email WINS EVERY SINGLE TIME." Replying to a critic, the owner added that "our spam rates are way low". That's their test and their spam rate, and we haven't seen the data behind either.

None of the five is clever. They're setup work, and that's most of the answer to how to automate email marketing for ecommerce: somebody has to build them once. Two operators handed email work to AI. A revenue lead at a DTC fashion brand said an AI model built the campaign strategy, wrote the emails, ran the revenue analysis and handled the email platform's operations, and claimed revenue rose 54% with a "team" of two AI tools and one person. An owner of several online stores said a browser agent audited every store, logged into every app they use and built pre-purchase and post-purchase flows better than they could have. That owner reported no result.

Two email specialists in our research read the attributed revenue figure differently, and only one of them still leads with it. A retention consultant who works for large DTC brands stopped treating the email platform's attributed revenue as a success metric about a year ago. The figure, in the consultant's words, "tells you that someone received an email and then purchased within a certain window. It doesn't tell you whether that person would've bought anyway." The consultant has seen brands celebrate a large email share when a big chunk of it came from returning customers who would have bought regardless. The email happened to be in the inbox, so the attribution model gave it the credit.

The auditor from the first table still starts with the share. The second check is how that revenue splits between campaigns and flows, where the "goal is close to 50/50". On the live account at 26%, the split was 62% campaigns and 37% flows. "So the flows are where the next money sits." The auditor's view of the headline figure is blunt: "The attributed total on its own tells you what email claimed."

They agree on more than it first appears. The consultant still tracks attributed revenue "directionally", and the auditor still opens with it. Neither trusts the total alone. Where they part is what goes next to it. The auditor stays inside the platform and reads the split. For the consultant it's customer numbers, because otherwise "you end up optimizing for a dashboard instead of for the business."

When email went wrong in the cases we found, the words weren't what broke. The trouble was getting to the inbox, the platform itself, and what it cost.

• An email specialist had an abandoned-checkout email that "initially landed in spam", although the design, the offer and the automation were all fine. The fix was to simplify the content, review the sending-domain authentication and adjust the sending approach. The specialist's warning for the holiday peak is that it "is not the time to suddenly reactivate every subscriber who has ignored the brand for years."
• The owner of the brand they call 7-8 figure, where email brings in 10% of revenue, delegated email to raise the open rate and "decrease spam percent". On another day the same owner was asking whether the email platform was "down for everyone too right now or just me?"
• A hardware brand's founder called the email platform "10x too expensive" and said they were still paying it $3,000 a month, "for now but not for long". A DTC brand operator called it the go-to choice but "quite expensive", and added, "I don't use 80% of the features".
• In one relayed story, an agency moved a client off its email platform weeks before the holiday peak, with $113K in email revenue making up 53% of the store's revenue: a channel carrying half the store, moved at the riskiest moment. We only have it secondhand, from someone arguing that the platform the client left was the real risk, so it's one side's telling.

Then there's the owner with the plain-text tests. Years earlier, that owner reported what they called their best campaign ever on the email platform, and it came from discontinuing half the items in the store. "50% of our SKUs accounted for 10% of our sales. No discount or sale, just FOMO." Set that against the cannabis account's welcome emails, which were mostly the discount.

Say a store owner doing about $60,000 a month comes to us with a popup, a list and one welcome email. We'd look first at which of the five flows exist, and in that setup four of them don't. Next is whether the sending domain is authenticated, and that gets settled before anyone sends more. Third, we'd want a repeat-purchase rate and revenue per subscriber over 90 days, built from the store's own orders rather than the platform's attribution. The same orders drive updates and support replies too, which is where automating the rest of the store, from orders to support comes in. Until that third number exists, the only score the store's email has is the one the platform gives itself.