Most automation advice for online stores is a list of fifty things you could do, sorted by nothing. You finish reading it more tired than when you started, and the tab stays open for three weeks until you close it out of guilt.
So here's a shorter list, and it's ordered. Five automations, in the sequence I'd install them for a store that's already selling. Each one fixes a specific leak you can probably name from memory, and each is something you install and point at your own accounts, not a project you build from scratch.
The order isn't arbitrary. It starts closest to the register and works outward, because the fastest way to trust an automation is to watch it make or save money in the first week.
1. The cart nudge that lifts the order before checkout
You've watched it happen in your own analytics. Someone fills a cart, lands a few dollars under your free-shipping threshold, and leaves the exact amount that would have pushed them over sitting on the table. Nothing on the page told them they were close.
The first automation to install is the one that does the telling: a live nudge that shows a shopper how far they are from free shipping and what one more item would do. It works on carts that are already in motion, which is why it pays back quickest. This is the AOV side of the cart. For the other side, recovering the ones who leave, see the abandoned-cart piece.
Install it: Cart free-shipping progress nudge.
2. The win-back that fires the moment someone cancels
The worst time to have no system is the second a subscriber clicks cancel. Right now, for most stores, that click is the end of the conversation. No pause offer, no smaller plan, no "before you go." The customer you spent real money to acquire walks out a door you left unlocked.
Second to install is the automated win-back: when someone hits cancel, they get an offer to pause or step down instead of leaving outright. Retained revenue compounds in a way a new sale doesn't, and this one runs on the moment you're currently sleeping through.
Install it: Save cancelling subscribers with a win-back offer.
3. The search terms with a return you're not acting on
If you run paid search, there are converting terms buried in your reports that never get added to a campaign. They show a strong return, they sit in a report that stays closed, and the account leaves money on the table every month they go unactioned. The strategy is fine. The time to action them by hand is what's missing.
Third is the automation that harvests those terms and surfaces them for action, so the same ad budget buys more conversions. It's the highest-return pick in this list for a store already spending on search.
Install it: Harvest paid search terms with a return sitting unadded.
4. The support agent for the same five questions
By now you've automated three things that touch revenue directly, so the next move is the biggest block of time you keep spending by hand. For most stores that's support, and most of support is the same handful of questions asked in different words. Where's my order, can I change my size, what's your return window.
Fourth is a trained agent that answers the repeat questions and hands the genuinely new ones to a person. It's the closest thing on this list to hiring, without the hire, and it buys back hours instead of dollars, which is why it comes after the revenue picks rather than before.
Install it: Deflect repeat support tickets with a trained AI agent.
5. The product feed that reads the way people search
Once the flows and the support load are handled, the highest-leverage thing left is usually the feed behind your Shopping and PMax campaigns. Feed titles that read the way your catalog is organized, rather than the way people actually search, quietly cap what those campaigns can do, and a campaign with no segmentation spends evenly across products that don't deserve it.
Fifth is the feed optimization that rewrites titles toward real search language and gives the campaign structure to work with. It's last because it pays off most once the earlier four are already running.
Install it: Merchant Center feed optimization for Shopping and PMax.
The pattern under all five
Read the list back and every pick does the same thing before it does anything clever: it stops you from losing something you already have. A cart already in flight. A subscriber already acquired. A search term already converting. A support hour already yours. A campaign already running. That's the whole first mile of store automation, and it's a better place to start than any list of fifty.
You don't have to build any of these. Each one exists as a kit you install and point at your own store, so someone else has already met the edge cases and you get the version that comes after their bad week.
I'm Kent Clark. I run shopautomations.ai, where we build working automations for online stores, one workflow at a time, and test every one before it ships. The full set for e-commerce is here, sorted by the work you actually do. If you're not sure which leak to close first, the AI ROI Diagnostic is free with a signup and hands you a scored shortlist for your store.