Is your Shopify app stack eating your margin?

Every Shopify app gets installed for a good reason. A review widget to lift conversion, an upsell app for average order value, a subscriptions tool, a back-in-stock alert, a loyalty program, a page builder for the summer campaign. Each one costs a few dollars a month, or a small cut of each order, and on its own none of them looks like a problem.

The problem is the stack. When we audit stores it's common to find a few dozen apps installed, a handful doing two jobs that overlap, a few nobody on the team can explain, and some left over from a campaign that ended two years ago. Together they cost real money every month and slow the store down, and almost nobody looks at that cost against what each order actually makes.

This is how we audit an app stack, and the one number we use to decide what stays.

1. Stop judging apps by their monthly fee

The usual app review goes down the billing page and asks "is $49 a month a lot?". That's the wrong question, because the fee only means something next to the orders it sits on.

A $300 a month app on a store doing 10,000 orders a month costs 3 cents an order. The same app on a store doing 600 orders a month costs 50 cents an order. If that smaller store makes $12 of margin on an average order, the app is eating about 4% of it before it has done anything.

So the first step is to turn every app into a cost per order:

App cost per order = (monthly fee + usage or per-order charges) ÷ monthly orders

Do that for every app on the billing page, including the ones billed through a separate account, and add them up. That total is your app stack's cost per order, and it belongs in the same conversation as shipping, payment fees and discounts.

2. Count the costs that don't show up on the invoice

The subscription fee is the visible part. Three other costs tend to be bigger.

Per-order and revenue-share fees. Plenty of apps for reviews, upsells, subscriptions and shipping protection charge a percentage of revenue or a fee per order on top of, or instead of, a flat plan. These grow with the store, so an app that was cheap at launch can be one of your bigger lines a year later. Pull the usage charges from the last three months, not the plan price.

Speed. Every app that loads scripts on the storefront adds weight, and slower pages convert worse, especially on mobile. Uninstalled apps often leave code behind in the theme too, so a store can carry the cost of apps it stopped paying for long ago. You can't put an exact dollar figure on this without testing, but a theme that loads scripts from a dozen apps on every page is worth a proper look.

Overlap. Two apps doing the same job is more common than you'd think: a review app and a loyalty app both collecting reviews, an upsell app and a bundle app both changing the cart, Shopify's own features now covering something you still pay a third party for. You pay twice and the scripts can fight each other.

3. Measure each app against contribution margin per order

Revenue is a poor test for an app, because an app can lift revenue and still lose money once you take out what it costs and what it gives away. An upsell app that adds $8 to the average order by discounting a second item 30% might be adding very little margin. A free-shipping-threshold bar might be lifting order value while pushing more orders into a shipping band that costs you more.

The test we use is contribution margin per order: what's left from an average order after product costs, shipping, payment fees, discounts and marketing. For each app that claims to drive sales, ask two questions.

1. What does it cost per order, using the number from step 1 plus any per-order fees?
2. What does it add to contribution margin per order, after the discounts or freebies it hands out?

If the second number isn't comfortably bigger than the first, the app is costing you money no matter what its dashboard says. App dashboards tend to report the revenue they "influenced", which usually counts orders that would have happened anyway. Where you can, test it properly: switch it off for two to four weeks, or run it on half your traffic, and compare margin per order, not just conversion rate.

Apps that don't touch sales (shipping labels, accounting connectors, back-office tools) get a simpler test: does it save more staff time than it costs, and is there something cheaper or already built into Shopify that does the same?

4. Run the audit in an afternoon

Here's the routine we take clients through.

1. Export the list. Every installed app, its plan, and the last three months of charges including usage fees. Check Shopify billing and any apps billed separately by card.
2. Name an owner. For each app, write down who uses it and what job it does. If nobody can answer, it goes on the removal list.
3. Work out cost per order. Average monthly cost divided by average monthly orders, as in step 1.
4. Sort into three groups. Apps that earn their place on margin, apps that do a necessary job (keep, but check for cheaper options), and apps that can't show either.
5. Check for overlap. Group apps by job (reviews, upsells, loyalty, subscriptions, search, page building) and look for doubles.
6. Remove properly. Uninstall, then clean the leftover code from the theme. Test checkout, product pages and the cart afterwards.
7. Schedule the next one. Once a quarter, or before a big trading period like Black Friday, when speed and margin both matter most.

A worked example

Take an illustrative store doing 2,000 orders a month at a $70 average order value, with $14 of contribution margin per order. It runs 28 apps.

The billing export shows $1,900 a month in flat fees plus about $1,100 in usage and revenue-share charges, so $3,000 a month, or $1.50 an order. That's more than 10% of the store's contribution margin per order going on apps.

The audit finds four apps nobody uses, two review tools doing the same job, and an upsell app whose discounts give away most of the extra order value it creates. Cutting those brings the stack to roughly $1,800 a month, or 90 cents an order. That's 60 cents back on every order, around $14,000 a year at the same volume, before any speed gain from lighter pages.

Your numbers will be different, but the pattern is common: a few apps carry the store, and a long tail costs more than it earns.

What to do next

Start with the billing export and the cost-per-order sum. Most teams have never seen their app stack as a cost per order, and that one number usually makes the next decisions easy. If you want help working through the list, or cleaning up a theme after years of installs and uninstalls, that's the kind of job we do for Shopify stores every week.

Book a free call and we'll start with your app list.

Frequently asked questions

Spotted something wrong or out of date? Email hello@kirkbyconsulting.com with the details and we will correct it.