The stack · updated 20 September 2026
18 of them, with what each one costs you and where it stops being worth it. We go back through this every quarter and things move between sections, usually because pricing changed or Shopify shipped the feature natively.
4 we use10 it depends4 we skip
The monthly bill
This is roughly what we find when we open a store that has been running a few years. Nobody sets out to spend it. Apps get added for a campaign and never removed, and two of them end up doing the same job.
18
apps weighed in the open
14
we build with, and the point each stops being worth it
4
we will not install, with the reasoning written out
These go on most builds. If we are not installing one of them, we will tell you why.
We use it
Still the first thing we wire after checkout works.
Flows, segmentation and deliverability that nothing at this price band matches. The Shopify integration is deep enough to segment on things you actually care about, like second purchase timing.
The catch The bill climbs steeply past 50k profiles. We will tell you when you have outgrown the pricing rather than letting it creep.
We use it
Reviews that do not cost you a second of load time.
Flat pricing, clean markup, proper review schema, and it does not drag a framework onto the product page to render five stars.
The catch The default widget styling is plain. Budget an hour to make it look like your store rather than like an app.
We use it
Free, native, and it replaces three apps you were about to install.
Order tagging, fraud holds, inventory alerts, customer tagging — all first-party, no extra script on the storefront, no monthly fee.
The catch It runs on a delay. If you need something to happen inside a second, Flow is the wrong tool.
We use it
The native answer to selling in more than one country.
Real multi-currency with settlement, domain-per-market, local payment methods and duties handled at checkout. Built in, so nothing extra loads.
The catch Translation is a separate problem — Markets handles money and routing, not language.
Good tools with a threshold. Below a certain size, each of these costs more than it brings back.
Depends
Worth it once support is somebody full-time job.
Order context sits next to the ticket, so agents refund and reship without tab-hopping. It genuinely pays back at volume.
The catch Below roughly thirty tickets a day a shared inbox does the same job for a fraction of the cost.
Depends
Only if subscriptions are the business model, not a feature.
The most mature subscription tooling on Shopify — dunning, swaps, prepaid plans, a customer portal that does not embarrass you.
The catch If subscriptions are six percent of revenue, you are paying for and maintaining a second checkout for six percent of revenue.
Depends
The lighter subscription option, built on native Shopify selling plans.
Because it uses selling plans rather than a parallel checkout, the buying experience stays inside Shopify and migrations later hurt less.
The catch Fewer edge-case features than Recharge. If your plan logic is complicated, check it handles your case before committing.
Depends
Upsells and cross-sells, once you have the traffic to test them.
Cart and post-purchase offers that can be tested properly rather than guessed at.
The catch Under a few thousand sessions a month you cannot reach significance, so you are just adding weight to the cart for a feeling.
Depends
Real search, for catalogues big enough to need it.
Typo tolerance, synonyms and merchandising rules that turn search into a revenue surface instead of a dead end.
The catch Under roughly five hundred SKUs, native search plus decent collection structure is enough. Spend the money on photography.
Depends
Branded tracking, for when where is my order is drowning support.
A tracking page on your own domain deflects the single most common support ticket there is.
The catch It only pays for itself at volume. Below that, a good shipping confirmation email does most of the work for free.
Depends
Loyalty works in repeat-purchase categories and nowhere else.
Points and referrals are straightforward to run, and the widget is lighter than most of the category.
The catch If people buy from you once every three years, a loyalty programme is a launcher icon nobody taps. Category first, app second.
Depends
Attribution for stores spending real money on ads.
Pulls spend and revenue into one view and models the gap that iOS left in platform reporting.
The catch Under roughly twenty thousand a month in ad spend, the subscription costs more than the decisions it improves.
Depends
Translation when you need it live next month, not next quarter.
Fast to deploy, handles app strings and checkout, and the machine baseline can be corrected by a human afterwards.
The catch It is a subscription forever. If you only need two languages and have the copy, native Shopify translations cost nothing.
Depends
Pre-orders, deposits, backorders and waitlists — controlled per variant.
One size can sit on pre-order while the rest ship normally, on the same product page. Almost nothing else in the category handles mixed states without splitting the product in two.
The catch If you never sell ahead of stock, you do not need it. We will say so.
The reasoning matters more than the verdict, so it is written out for each one.
Won't install
An app that adds a script to your store, to make your store faster.
Most of them do lazy-loading and minification that a decent theme already does, then add their own runtime on top and ask for a monthly fee. The ones that genuinely help are usually solving a problem created by another app you could remove instead.
The catch If a store is slow, we will show you which scripts are costing you the time before anyone suggests installing anything.
Won't install
Selling you back what Shopify and a competent theme already do.
Meta titles, alt text, sitemaps and product structured data are theme and platform work. An app that writes them for a monthly fee is renting you something you own.
The catch If your SEO genuinely is broken, the problem is usually site structure, internal linking or thin collection pages — none of which an app fixes.
Won't install
Superseded by Shopify Markets, and worse than it.
They change the number on the page without changing the currency at checkout, so shoppers are quoted one price and charged another. That is a refund and a chargeback waiting to happen.
The catch Markets does this natively with real settlement. If you are running a converter app, removing it is usually a same-day job.
Won't install
A third-party script, on every page, to show pictures you already own.
They load an external bundle and pull images from an API you do not control, so the section breaks whenever that API changes or the token expires. It is also the slowest way to put six photographs on a page.
The catch If the social proof is worth having, we will build it as a theme section with your own images, and it will keep working.
This is where we start from, not what you end up with. What you sell, how often people come back and how many SKUs you carry all move things around. Working that out is most of the first conversation.