Retention is the easiest ecommerce service to sell once you are in the room. Every merchant already suspects their repeat rate is worse than it should be, and the arithmetic of an owned channel is not controversial. The hard part is getting into the room. A slow site announces itself. A thin catalog announces itself. A missing welcome flow leaves almost no public trace — almost being the operative word.
The app stack is the only public trace retention leaves
Every store record carries an apps field: the detected application and technology stack. Email service providers, popup and capture builders, review widgets, loyalty programmes, subscription managers, SMS platforms, referral tools. These leave fingerprints in the page, and those fingerprints are the closest thing to a public statement about how a merchant thinks.
You are not reading revenue. You are reading intent. A merchant who installed an email platform decided at some point that email mattered. A merchant who never did either has not had the thought or had it and shelved it. Those are two different conversations, two different sequences, and two different price points — and telling them apart before you write is most of the work.
As always, detection is a probability rather than a proof. Some merchants send from a platform-native tool that leaves nothing detectable, some run everything server-side. So phrase the observation as a question. “I could not find an email capture on the site — am I missing it, or is that genuinely not running yet?” costs you nothing when you are right and rescues the email when you are wrong.
Gap A: real traffic, no tooling whatsoever
The cleanest pattern in the set. A store with genuine monthly traffic — Semrush worldwide figures cover 118,324 of the stores in the database — a catalog with real depth, and nothing in the detected stack beyond platform defaults and a payment gateway.
This merchant is earning or buying traffic and then discarding it. Every visitor who does not convert on the first session leaves with nothing attached to them. Every visitor who does convert never hears from the brand again except through a shipping notification written by the platform in 2019. The leak is not subtle and it is not seasonal; it runs every single day.
The pitch is arithmetic rather than opinion, which is why it converts. Take their monthly traffic, apply a conservative capture rate for a competently placed signup incentive, apply a conservative flow conversion rate, and multiply by the price points visible on their own product pages. Present it as a reasoned range and label it as an estimate — never as a forecast, never as a guarantee. Most merchants have never seen that number written down anywhere, and the reflex is to ask where it came from. That question is your meeting.
Gap B: the platform is installed and nothing is running
More common than gap A, more irritating for the merchant, and therefore a better sale. The stack shows an email platform. The merchant pays for it every month. What it actually does is send an occasional campaign blast when somebody on the team remembers, and fire whatever default flow shipped in the onboarding wizard.
You can confirm this from outside in ten minutes without contacting anybody. Subscribe with a throwaway address. Add something to the cart on a second visit. Abandon it. Wait seventy-two hours and record what arrives. Usually the answer is one generic message and then silence. At that point you know more about their email programme than the person paying for it.
That test is also your artifact. A single page showing exactly what a new subscriber received across four days, next to what a well-run store in their category would have sent, is more persuasive than any deck you could build. It costs twenty minutes per prospect, so reserve it for the thirty or forty stores at the top of your list rather than spraying it across four hundred.
Gap C: a catalog shape that makes retention structurally obvious
Product count, category, visible price points and currency together describe whether a business is structurally a repeat-purchase business. A surprising number of merchants run replenishment catalogs with acquisition-only marketing, which is the most expensive mistake available in ecommerce.
Consumables carry a clock inside the product. Supplements, coffee, skincare, pet food, cartridges, filters, cosmetics. If the stack shows no subscription tooling and no email platform, the merchant is paying full acquisition cost to re-win the same customer every cycle. You do not need to see their accounts to know that hurts; you need to see a category and a product count.
The opposite catalog — furniture, high-ticket equipment, made-to-order, anything bought once a decade — is not a replenishment play at all. It is a referral, review-generation and cross-sell play, with a completely different flow architecture and a completely different first email. Pitching a replenishment programme to a furniture merchant marks you instantly as somebody who did not look before writing.
Gap D: the platform sets the default behaviour
WooCommerce 58,549, Shopify 50,002, PrestaShop 16,501, JouwWeb 15,477, Odoo 4,510, Squarespace 2,447, Magento 1,761, and sixty distinct platforms in total. Platform is not only a technical field. It is a behavioural one, because each ecosystem trains its merchants differently.
Shopify merchants live inside an app marketplace that pushes retention tooling at them relentlessly. When a Shopify store has no email platform detected, the merchant has actively declined at least once, and your email has to overcome a prior decision. When one does have a platform installed, the flows probably exist in some skeletal default state, and your offer is optimisation rather than installation — smaller opening ticket, faster proof, easier upsell.
WooCommerce and PrestaShop merchants are usually running transactional mail through a plugin and nothing beyond it. No marketplace is nagging them, no onboarding wizard walked them through a welcome series. The gap is wider, the merchant is less aware it exists, and the first conversation is more educational. That means a longer cycle and materially less competition in the inbox, which often produces a better retainer because you end up owning the entire channel rather than auditing somebody else’s.
Site-builder platforms skew towards smaller operations with thinner catalogs. Not worthless, but price the engagement accordingly and expect volume rather than depth. Knowing which of the sixty platforms you are willing to serve is a positioning decision, and making it explicitly beats making it accidentally.
The order to pitch the flows in
Never lead with a full lifecycle programme. It sounds expensive, it sounds slow, and it forces the merchant to imagine a six-month project before they have seen you deliver anything. Lead with the two flows that pay for the engagement inside the first month, then expand from a position of proof.
Welcome comes first. A subscriber who has just handed over an address is at peak intent, and most stores answer that moment with a single discount code and nothing else. A properly built welcome sequence runs eight to ten messages across two or three weeks — brand story, social proof, category education, objection handling, best-seller spotlight — with a first-purchase incentive carried in the opening message rather than dribbled out later.
Cart and browse abandonment come second, because they recover revenue the merchant has already spent money to earn. Render the actual cart contents inside the email so returning takes one click and no thought. Post-purchase comes third: somebody who has just entered card details is the warmest audience the brand will ever address, and almost every store answers that moment with a tracking link and nothing else. Winback and replenishment come last, timed to the real consumption cycle of the product rather than to a round number of days somebody picked in a settings panel.
One operational discipline worth handing the client on day one: before any large campaign send, cut the list to people who opened or clicked in the last ninety days. Blasting the full list protects nothing and slowly destroys the sending reputation the entire programme depends on. Merchants who learn that rule from you tend to keep paying you.
Turning four gaps into four sequences
The mistake is writing one retention email and aiming it at everybody. Four gaps means four opening observations, four implications, four artifacts. The body of your offer can be identical; the first eighty words cannot be.
The query shape is straightforward through the REST API or the MCP server: platform, country, a monthly traffic band, a product count band, and an apps filter that includes or excludes specific tooling. Segment first, write second. A list of four hundred stores split into four groups of one hundred outperforms four hundred identical sends by a wide margin, and takes about the same amount of time.
Then keep the first email to one observation, one implication, one artifact and one question. Something like: you are doing somewhere around this much traffic a month and I could not find an email capture anywhere on the site; I ran conservative numbers on what a welcome flow would be worth to you; it is one page, want it? Nothing in that email is unverifiable, and nothing in it requires a call.
Underneath all of it, remember what you are actually selling. Not emails. The difference between a store that rents its customers from an ad platform and one that owns them outright. That framing survives a price objection. A feature list does not.