Search for app recommendations and you will find the same twenty names ranked by affiliate commission. That is not a useful way to build a stack. A store doing 40,000 dollars a month with three well-chosen apps will beat one running eighteen, and the second store will be slower, more expensive to run, and harder to change. The question worth asking is not which app is best. It is which categories deserve a slot at all, and what each one has to earn to keep it.
The best Shopify apps for your store are the ones tied to a job you can name and a number you can watch. Judge by category first, reviews, email, upsell, search, subscriptions, analytics, then pick the lightest option that does the job. Every app carries a monthly fee and a speed cost, so each one needs to pay for both.
Why app listicles lead you wrong
Ranked lists optimise for breadth, not for your store. They cannot know your margin, your order volume, your team's capacity or how much custom work already sits in your theme. Worse, they encourage a shopping mentality: install first, find the use later. That is how stores end up with two review platforms, three analytics scripts and a pop-up tool nobody has opened since the person who set it up left.
A better starting point is subtraction. Assume the answer is no app, then require a specific case for each one. That case has three parts: the job it does, the number that should move, and who owns it. If any of those three is missing, the app is a maybe, and maybes should not be installed.
Which app categories usually earn a slot?
Most healthy stacks for a US small business sit somewhere between four and eight apps. The categories below cover what we see paying off most often. Pricing is a general market range, not a quote, and it moves with order volume, contact list size and feature tier.
| Category | Job it does | Typical monthly range | Earns its place when |
|---|---|---|---|
| Reviews and UGC | Collects, displays and syndicates customer reviews and photos | Free tier to 150 dollars | You sell considered purchases where social proof changes the decision |
| Email and SMS | Flows, campaigns, segmentation, abandoned cart and post-purchase | 20 to 500 dollars, scales with list size | Almost always. This is the one category most stores under-invest in |
| Upsell and bundles | Cart add-ons, product bundles, post-purchase offers | 20 to 200 dollars | Your catalogue has genuine complements and your AOV has room to move |
| Site search and filtering | Better on-site search, synonyms, faceted collection filtering | Free tier to 300 dollars | You have more than roughly 150 SKUs or heavy attribute-based browsing |
| Subscriptions | Recurring billing, customer portal, dunning | Flat fee plus a percentage of subscription revenue | Replenishment is a real buying pattern for your product, not a hope |
| Analytics and attribution | Channel attribution, funnel reporting, profit tracking | 50 to 400 dollars | You spend enough on ads that mis-attribution costs more than the tool |
| Shipping and returns | Rate logic, tracking pages, self-serve returns | Free tier to 200 dollars | Returns volume is eating support hours or shipping rules exceed native logic |
| Loyalty and referral | Points, tiers, referral mechanics | 50 to 300 dollars | You already have repeat buyers to reward, rather than hoping to create them |
Two categories deserve scepticism. Loyalty programmes are frequently installed by stores with no repeat purchase behaviour, where the points widget becomes decoration. And attribution tools are often bought to settle an argument about ad spend that better tracking hygiene would have resolved for free.
Before adding any app, write one sentence: this app exists to move [metric] from [current] to [target] by [date], owned by [person]. If you cannot fill in the blanks with real numbers, you are not ready to install it.
What do apps actually cost?
The subscription fee is the visible part. Two other costs matter more over a year.
The first is revenue share. Subscription platforms, some upsell tools and most loyalty products take a percentage of the revenue they touch. At low volume that is cheap. At scale it can quietly become one of your larger line items, and it is rarely renegotiated because nobody re-reads the pricing page after signing up.
The second is operational cost. Every app is a system someone has to configure, monitor and eventually migrate off. A tool that saves five hours a month but needs three hours of maintenance is not the win it looked like in the demo. Migration is the hidden one: review content, subscriber lists and loyalty balances are all painful to move, and some vendors make it deliberately harder.
Add it up honestly. Eight apps averaging 60 dollars a month is 5,760 dollars a year before revenue share, which is real money for a small business and often more than the custom development that would have replaced two of them permanently.
How much do apps slow a store down?
This is where the damage tends to be invisible until it is not. Most Shopify apps inject scripts, styles or theme blocks into your storefront. Each one adds requests, JavaScript to parse, and often a layout shift when the injected content arrives late. A single app rarely wrecks performance. Six of them, each adding a modest amount, do it together.
What makes it worse is residue. Uninstalling an app frequently leaves code behind in your theme, particularly if it was added before Shopify's app embed model or if a developer hand-placed the snippet. We regularly open theme files and find script tags for tools the client stopped paying for a year ago.
- Run a performance test on a product page and a collection page, on mobile, and save the result as a baseline.
- Look at the request waterfall and identify third-party domains. Match each one to an app you know you are paying for.
- Flag anything you cannot identify. That is usually residue or a tool nobody owns.
- Remove one app at a time, retest, and record the difference. One at a time matters, because batching removals means you learn nothing about which one mattered.
- Check your theme files for orphaned snippets after each uninstall rather than trusting the removal process.
If page weight is already a known problem, treat app reduction as part of the same workstream as any other site build or rebuild work rather than a separate cleanup you will get to later.
How do you audit a stack you inherited?
Most stores we look at have never done this. It takes an afternoon and usually pays for itself in the first month.
Start with the billing page rather than the app list, because billing tells you the truth about what you are paying for including tools that no longer appear anywhere in your admin. Then for each app, answer four questions: what job does it do, who logs into it, when did anyone last change a setting, and what breaks if we remove it tomorrow.
Sort the results into four buckets. Keep, where the job is real and the tool is working. Downgrade, where you are on a tier you have outgrown in the wrong direction. Replace, where two apps overlap or where native Shopify functionality has caught up since you installed. And remove, where nobody can name the job.
The replace bucket is the interesting one. Shopify's native features have absorbed a lot of what used to require apps, particularly around discounts, bundles, markets and basic customer accounts. Stores rarely revisit decisions made three years ago, so it is common to find two or three apps paying rent for something the platform now does.
Set a calendar reminder for a 60 minute app review every quarter. Cancel anything with no owner and no setting change in six months. The discipline matters more than the individual decision.
When is custom development the better answer?
Apps are the right call when the job is standard, the maintenance burden is real, and the vendor is doing something genuinely hard such as deliverability, payment compliance or search relevance. You do not want to build your own email sending infrastructure.
Custom becomes better when the app is doing something simple, when you are paying a percentage of revenue for logic you could own, or when the app's constraints are forcing your merchandising decisions. A bundle feature that costs 200 dollars a month and cannot express your actual bundle rules is worse than a few days of development. As a rough guide, if an app costs more than 2,000 dollars a year and does one narrow thing, get a development estimate before renewing.
The trap is building custom versions of things that need ongoing vendor attention. Search relevance, tax logic and fraud scoring all get worse over time if nobody maintains them, and most small businesses do not have anyone to maintain them.
What should you do first?
Cut before you add. Almost every store we audit can remove two apps without losing anything, and the speed and clarity gained make the next decision easier. Then fill genuine gaps in order of impact: email and SMS first if that is thin, then reviews if you sell anything people research, then search if your catalogue is large enough to need it.
Resist the urge to solve a traffic problem with an on-site app. If not enough people are arriving, an upsell widget will not fix it, and that is a performance marketing conversation instead. If plenty arrive and few buy, the answer is usually in your pages before it is in your app list, which is where conversion work starts.
Frequently asked questions
How many Shopify apps is too many?
There is no hard number, though most healthy small business stacks land between four and eight. Count is the wrong test anyway. The real warning signs are apps nobody owns, two tools doing overlapping jobs, and third-party scripts on your storefront you cannot match to anything you pay for. A store with ten well-owned apps each tied to a metric is in better shape than one with five that nobody has logged into since setup.
Do Shopify apps really slow down a store?
Yes, though the damage is cumulative rather than dramatic. Apps that inject storefront scripts add network requests, JavaScript the browser must parse, and frequently a layout shift when their content arrives late. One app is usually fine. Six modest ones together are not. The bigger problem is residue: uninstalling an app often leaves snippets in the theme, so stores end up carrying code for tools they stopped paying for long ago.
Should I use free apps or paid ones?
Free tiers are a reasonable way to test whether a category earns a slot before committing budget. What they are not is a long-term plan for anything customer-facing. Free plans commonly cap the features that make the tool worthwhile, add vendor branding to your storefront, or limit volume in ways you will hit right when things are going well. Decide whether the job matters first, then pick the tier that does it properly.
What happens to my data if I uninstall an app?
It depends entirely on the vendor, and the difference is worth checking before you sign up rather than when you leave. Review content, email subscriber lists, loyalty point balances and subscription contracts are the categories where migration genuinely hurts. Ask what the export format is and whether it can be re-imported elsewhere. Vendors with a clean answer tend to have a clean product. Vendors who get vague are telling you something.
When should I build custom instead of installing an app?
Look at custom when the app does one narrow thing, when it takes a percentage of revenue for logic you could own outright, or when its limits are dictating how you merchandise. A rough trigger: any app costing more than 2,000 dollars a year for a single feature deserves a development estimate before you renew. The exception is anything that decays without ongoing vendor attention, such as deliverability, search relevance, tax logic or fraud scoring.
How often should I review my app stack?
Quarterly works for most stores, and an hour is usually enough. Start from your billing page rather than the installed app list, because billing surfaces charges that no longer correspond to anything you can see in the admin. For each line, ask who owns it and when a setting last changed. Also check whether native Shopify functionality has caught up, since discounts, bundles and customer accounts have all absorbed jobs that once needed a separate app.
The takeaway
Treat every app as a tenant paying rent in money, speed and attention. Most stores can cut two today and lose nothing. Do that first, then fill the gaps that matter with the lightest tool that does the job, and write down who owns each one. Review the list quarterly from the billing page. That habit will serve you better than any ranked list of app names ever will.