Marketing automation multiplies whatever your store already does. If it converts at 1.2% on mobile, you're automating 1.2%. The order you buy in matters more than which tools you pick.
Search "ecommerce marketing tools" and you get lists of forty products. Email tools, SMS tools, popup tools, loyalty tools, review tools, affiliate tools, each with three bullet points and a star rating nobody computed. Every entry is recommended. None of them are compared.
That's not a shortlist. That's a directory.
The useful question isn't which tool. It's what you're automating, and whether the thing underneath the automation is worth multiplying yet.
Marketing automation multiplies. It doesn't create.
Here's the part the listicles skip.
Every marketing tool you buy takes a number your store already produces and moves more people through it. An email platform sends a flow to people who gave you an address. An ads platform buys sessions. A popup captures a percentage of visitors. All of them end at the same place, which is your storefront, converting at whatever rate it converts at.
Median session conversion on Shopify sits around 1.9% on desktop and 1.2% on mobile. Those are medians across thousands of stores, and mobile is where most of your traffic is.
So when you automate, you're automating 1.2%.
Buy the email platform first and you'll send beautifully designed messages to a product page that doesn't answer the question blocking the sale. The tool works. The flow fires. The revenue doesn't move much, and the tool gets blamed for it.
At Qosmic we run a 500-check audit across a store's pages, on mobile and desktop, and the thing that shows up over and over is stores paying for sophisticated acquisition into pages that fail basic checks. Nobody set out to do that. It happens because tools get bought in the order they get marketed to you, and messaging tools market hardest.
The three jobs, and which are actually solved
Strip the category down and marketing automation does three things.
It brings people in. Paid ads, organic search, affiliates, influencers. Tools that spend money or find audiences.
It talks to people. Email, SMS, push, retargeting. Tools that move someone from "knows you exist" to "on the site again."
It changes what people see. The store itself. Which products surface, what the page says, what the offer is, what happens at the cart.
The first two are crowded with mature products and you should buy from them. The third is where the automation mostly stops, and it's the one that decides whether the other two paid for themselves.
What to buy, in order
Messaging: buy this, and buy it properly
Klaviyo is the default on Shopify for a reason. Omnisend is cheaper and good if you want email and SMS in one place without Klaviyo's pricing curve. Postscript and Attentive are the serious SMS options once SMS is a real channel for you rather than an experiment.
We don't compete here and we don't want to. If you need flows, segmentation and a sending reputation, those are the right places to run that, and building it yourself is a bad use of a year.
Owned channels are worth real money. Email and SMS together account for roughly 28% of ecommerce revenue across verticals, and it runs higher in food, beverage and supplements where people reorder.
Read that number the other way around and you get the point of this article. About seven-tenths of revenue doesn't come from the messaging layer. It comes from people arriving at the store and deciding.
Measurement: buy one, not four
Triple Whale and Northbeam both do attribution for Shopify brands and both are fine. GA4 is free and most stores use maybe 10% of it.
The trap here isn't the tool, it's buying three of them and getting three different revenue numbers. Shopify counts an order when the order is placed. GA4 counts a session when a session starts. Those two never agree and nobody tells you why, so teams spend a quarter arguing about which dashboard is lying.
Pick one as the source of truth for revenue. Use the others for what they're actually good at.
The store itself: this is the gap
Here's the category that lists leave out, because for a long time there wasn't much to put in it.
Tools that tell you something is wrong are everywhere. Heatmaps show you where people click. Session recordings show you someone rage-clicking a dropdown. Analytics shows you a drop-off between the product page and the cart. Every one of them ends at a chart and a feeling.
What almost nothing does is make the change.
That's the job Qosmic took. The audit finds the opportunity, and then we build the variant and adjudicate it against real revenue, both halves. Not a recommendation you hand to a developer and revisit in six weeks. The change, built, and a verdict on whether it made money.
You can get to the same place without us. Hire an agency, or put a developer on it with a disciplined testing calendar. Plenty of good operators do exactly that and it works. It costs more and it's slower, and if you have the budget it's a legitimate choice.
What this looks like by stage
Under about $50k a month. One messaging platform, GA4, and fix the store. Conversion is flat across revenue bands up to roughly $50M, which means scale won't rescue a page that doesn't convert. You can't grow into it. The storefront and your own list are what you have to work with.
$50k to $500k a month. Add attribution, because you're now spending enough on ads that guessing costs real money. Start testing on the store rather than redesigning by opinion.
Above that. You have a team, and the constraint changes from tools to throughput. The question stops being what to buy and becomes how many changes you can get live and judged per month.
The tools nobody needs yet
A short list of things that get bought early and sit unused.
Loyalty platforms, before you have repeat customers to be loyal. Retention tooling on a store with no returning cohort is a monthly fee for a widget.
Personalization engines, before you have enough traffic for the segments to mean anything. Splitting 4,000 monthly sessions into six personas gives you six samples too small to act on.
A second popup tool, because the first one was configured badly. It wasn't the tool.
Anything with "AI" in the name that produces a report. The report isn't the work. You already have reports.
The order, condensed
- Fix what the store does per visitor. This multiplies everything after it.
- Buy one messaging platform and build the core flows properly.
- Pick one source of truth for revenue and stop arguing.
- Add attribution when ad spend makes guessing expensive.
- Then keep changing the store, permanently, and judge each change against revenue.
Step one and step five are the same work. That's not a coincidence. It's the only part of this list that compounds, and it's the part most stores treat as a project with an end date.
Cart abandonment sits between 70% and 74% depending on whose meta-analysis you read, and it has barely moved in a decade of better tools. All that automation, all those flows, and seven in ten carts still don't convert. That number isn't a messaging problem.
