The stack that works
at $5M
breaks at $20M.

I was advising an $8M annual revenue brand last year. Sharp team, solid product, growing fast. When I asked to see their Shopify billing, we found $14,200 a month going out the door on apps. Most of that stack had been assembled over two years without a single structured review. Four apps were doing overlapping work in the loyalty and review space. Two were on auto-pay for features the team had stopped using eight months earlier after switching to something else. Nobody had noticed because the credit card charge was buried in operating costs and the monthly amount, viewed in isolation, didn't feel alarming.

We ran a proper audit. The redundant and genuinely unused spend came to $4,800 a month. That's $57,600 a year. Not going to product. Not going to acquisition. Going to apps that nobody was actively choosing to pay for — they just hadn't gotten around to canceling them. This kind of cost drag compounds especially hard if you're also carrying working capital debt at 20%+ APR — the app bloat is directly reducing the margin that should be servicing that debt.

The stack conversation isn't primarily about what to add. It's about what to add, what to cut, and when each of those decisions changes as you move from one revenue tier to the next. That's what this post covers.

Most brands don't manage
their app spend.
They just accumulate it.

The average Shopify app costs between $58 and $67 a month. That sounds trivial. A brand running 20 apps is spending $1,200 to $1,400 a month before counting the major platforms — Klaviyo, Triple Whale, Gorgias, Loop Returns — which are each $200 to $1,200 a month depending on your volume. The combined number gets real fast, and unlike paid media or headcount, app spend has no natural forcing function to review it. You just keep getting charged.

App spending scales with revenue, roughly at 0.5–1.5% of net revenue for most Shopify brands. At $1M that's $5,000–$15,000 a year — manageable, easy to ignore. At $10M it's $50,000–$150,000 a year. At that level, undermanaged app spend is a meaningful drag on contribution margin, and it compounds because bloated stacks also slow your site down. Each app adds an average of 30–100ms to page load time. A 15-app stack has likely added a second or more to your load times, which hits conversion rate directly.

Stack decisions are also not symmetric by stage. The app that's worth $500/month at $3M revenue — because you have neither the team nor the process to do the thing manually — may be redundant at $15M when you've hired someone whose job description includes that function. Apps should fill gaps in capability that your team can't or shouldn't fill directly. When the team grows into a capability, the app covering it becomes a candidate for removal.

Five categories that
every serious Shopify brand
eventually needs.

Before the tier breakdown, there are five categories where every brand above a certain scale needs a real solution. These aren't aspirational — they're the table-stakes infrastructure of a functional Shopify brand. The debate at each tier is about which tool within each category, not whether the category is necessary.

Klaviyo deserves a note here because it's genuinely platform-layer, not just another app. Its Shopify integration is deep enough that flows like post-purchase sequences, winback campaigns, and abandoned cart recover revenue that would otherwise be gone. Brands that treat Klaviyo as just an email sender are leaving most of its value untouched. At $1M and above, Klaviyo is the retention operating system — the investment that makes every other retention-adjacent app more effective.

Lean, deliberate,
and hard to justify
adding to.

At this stage, simplicity is the strategy. You're still validating what works — which products convert, which acquisition channels are viable, what your actual repeat purchase rate is. The last thing you need is a complex stack that creates maintenance overhead and monthly costs before you've confirmed the model.

The $0–$2M Recommended Stack

Shopify Basic or Shopify plan (not Plus — Shopify Plus at $2,500/month is for brands above $1M in monthly revenue, not $1M in annual revenue). Shopify Payments built-in, which avoids the transaction fee on third-party gateways. Klaviyo, which is free up to 250 contacts — there's no reason not to start building flows from day one even at tiny list sizes, because the sequences you build early will compound. Judge.me Reviews at around $15/month — start here before Okendo or Yotpo, which charge a meaningful premium for features you don't need yet.

For SMS, Klaviyo SMS is the easiest path if you're already on Klaviyo — one platform, one integration, one billing relationship. Postscript is an alternative if you want a dedicated SMS tool with more advanced segmentation from the start. At this stage the difference matters less than the fact of having SMS at all. TikTok and Meta channels via Shopify's native integrations cover your pixel and catalog needs without an additional app.

That's the stack. Total app spend target: under $200/month.

This is where stack decisions
start having real
ROI implications.

At $2M–$5M, you have enough revenue to justify more sophisticated tooling, enough data to make decisions from, and enough repeat customers that retention infrastructure starts paying for itself. This is the tier where the right additions compound and the wrong ones just add cost.

What to Add at $2M–$5M

Upgrade reviews from Judge.me to Okendo or Yotpo. The step-up isn't just about better-looking widgets — it's about UGC collection, Q&A, deeper Klaviyo integration for post-review flows, and the ability to segment your best reviewers for loyalty and advocacy programs. Okendo starts around $88/month. That cost is justified when reviews are actively pulling conversion rate up and feeding your retention program.

Add a loyalty program: Smile.io or LoyaltyLion. Loyalty only matters when you have a meaningful base of repeat buyers to reward. Below $2M, the audience is often too small for a points program to change behavior at scale. Above $2M in most categories, you have enough repeat buyers that a real program — with VIP tiers, rewards on review submissions, and birthday campaigns — starts changing purchase frequency. Smile.io runs from $49/month (Starter) to $199/month (Growth). LoyaltyLion is more expensive and more customizable — right for brands that need complex tier logic.

Add retention analytics. Lifetimely or Triple Whale both give you contribution margin-based LTV tracking, which is the number you need to be making acquisition decisions from at this stage. Blended ROAS from Meta's reporting is not enough once you're spending real money. You need to know what each acquisition cohort actually contributes after COGS, returns, and shipping are factored in. Triple Whale runs $149/month at the entry tier — that's one of the clearest investments in the stack at this stage. For a full breakdown of how to build this model correctly, see the guide on the LTV math most Shopify brands are getting wrong.

Add Gorgias for customer service. The ROI on agent efficiency is real above 50 tickets a day. Gorgias pulls Shopify order data directly into the ticket view, so your support team doesn't have to tab-switch to answer most questions. Response times drop, ticket close rates improve, and agents handle more volume with the same headcount. Pricing starts around $60/month and scales with ticket volume.

Consider Loop Returns if you're in a high-return category. Apparel, footwear, anything with sizing complexity. Loop's exchange-over-refund flow regularly protects 15–25% of revenue that would otherwise have been refunded. The math works quickly if your return rate is above 15%.

Total app spend target at $2M–$5M: $800–$2,000/month.

The efficiency layer —
and where bloat
gets expensive fast.

The $5M–$20M tier is where the most consequential stack decisions happen, and where I see the most mistakes. Revenue is there to justify more sophisticated tooling. But this is also where brands accumulate apps from the previous two stages without auditing them, add new capabilities on top of old ones, and end up with a stack doing redundant work across five categories simultaneously. The stack changes at this stage are inseparable from the broader operational changes that define the $5M inflection — team structure, financial visibility, and retention economics all evolve at the same time.

What to Add at $5M–$20M

Attribution: Triple Whale or Northbeam — properly configured, not just installed. At $5M+ in paid spend, attribution accuracy is worth real money. The difference between a blended ROAS number from Meta's self-reported data and an actual contribution-margin-weighted attribution model is not a rounding error — it can point to completely different channel allocation decisions. Triple Whale runs $149–$219/month at the standard tier, which is one of the better returns in the stack at this revenue level. Northbeam is more expensive ($1,000/month and up) and targets brands with more complex multi-channel spend; evaluate it when you're spending seriously across Meta, Google, TikTok, and influencer simultaneously.

Subscriptions, if applicable: Recharge or Skio. Recharge is the long-established standard; Skio is newer with a better checkout experience and no-code migration tools. If subscriptions are part of your model and you've been managing them manually or through a lightweight tool, this is the tier where the right infrastructure pays for itself in subscription retention and payment recovery alone.

CRO tooling: Intelligems for price testing, VWO for broader A/B testing. At $5M+ in revenue, a 2% lift in conversion rate on your homepage or PDP is real money. You need a tool that can run tests cleanly without developer involvement on every experiment. Intelligems is specifically strong for Shopify price and offer testing — worth evaluating if pricing decisions are still being made without data. Before investing in testing infrastructure, run a manual product page audit first — the quick wins are usually visible without a test.

Inventory forecasting: Inventory Planner or Cogsy. Spreadsheet-based inventory management breaks above $5M when you're managing multiple SKUs, seasonal demand curves, and 3PL lead times simultaneously. A stockout on a top SKU during peak costs more than a year of Inventory Planner. A meaningful overstock ties up cash that should be in acquisition or operations. Both tools start around $99–$200/month and connect directly to Shopify.

SMS at scale: Postscript or Attentive. If you're still on Klaviyo SMS at $5M–$10M+ and sending volume is high, evaluate whether a dedicated SMS platform's deliverability, segmentation depth, and compliance tooling is worth the platform separation. Attentive targets brands at scale — pricing is volume-based and requires a conversation, not a self-serve signup. Postscript is more accessible and still strong at this tier.

Checkout extensibility (requires Shopify Plus): post-purchase upsell via Reconvert or ReCart, checkout extensions. The native Shopify checkout is good. Shopify Plus checkout with well-configured extensions is better — and the post-purchase upsell flow, when it's relevant to your product category, can add 5–15% incremental AOV without any additional acquisition spend. This requires upgrading to Shopify Plus ($2,500/month), which the math should support at $5M+ in annual revenue.

Total app spend target at $5M–$20M: $2,500–$6,000/month (including Shopify Plus).

At $20M+, the question
shifts from which apps
to how they integrate.

Something changes structurally at $20M. The conversation stops being about finding the right app for a function and starts being about how your core systems share data with each other. A fragmented stack with 40 apps and weak integrations is measurably worse than a tighter stack of 15 apps connected by clean data pipelines. The operational overhead of managing 40 semi-integrated tools, triaging support tickets from multiple vendors, and managing data discrepancies across systems is a real cost that doesn't show up in the monthly billing line.

What Changes at $20M–$50M

ERP integration becomes serious. NetSuite, Cin7, or Brightpearl with a Shopify connector. At this revenue level, managing inventory, purchase orders, financials, and fulfillment in Shopify alone is not viable. You need a system of record that can handle multi-location inventory, purchase order management, landed cost tracking, and clean financial reporting. The Shopify-to-ERP integration is one of the more technically complex things to set up correctly — budget for a proper implementation, not a weekend project.

Data warehouse and BI layer. Glew and Daasity are the Shopify-native options at this tier; Looker or Tableau with a custom data pipeline is the right answer for brands that have serious analytical ambitions or internal data teams. The goal: a single source of truth that pulls from Shopify, your ERP, your ad platforms, and your email/SMS platform. If you're making revenue, margin, and channel decisions from four different dashboards that don't agree with each other, you have a data architecture problem.

Shopify Flow for automation. At $5M, Shopify Flow is a nice-to-have. At $20M+, it's essential. Automating inventory tagging, fraud holds, loyalty tier updates, customer segmentation changes, and fulfillment routing saves meaningful team hours weekly and removes human error from repetitive decisions. The investment is team time to configure and maintain flows, not incremental software cost.

Customer data platform consideration: Klaviyo CDP or Segment. Klaviyo's built-in CDP capabilities cover most use cases if you're a pure-Shopify operation. Segment is the right answer when you have multiple data sources feeding customer profiles and need a central layer to manage identity resolution and data routing. This is a $20M+ decision for most brands — below that level, it's complexity without corresponding benefit.

3PL integration via ShipStation, ShipBob direct API, or WMS connection. If you're running your own warehouse or working with a large 3PL at scale, the Shopify connection to your fulfillment operation needs to be clean and automated. Manual order management at this volume is an error factory.

Dedicated fraud prevention: NoFraud or Signifyd. Shopify's native fraud tools are adequate below $5M in annual revenue. Above that threshold — and especially in high-AOV categories — dedicated fraud prevention with chargeback guarantees starts paying for itself. NoFraud and Signifyd both offer chargeback guarantees on approved orders, which changes the risk calculation materially.

Total app spend target at $20M–$50M: $6,000–$15,000/month for the app and platform layer. ERP and WMS costs at this scale often dwarf the app costs and should be budgeted separately.

The 45-minute audit
that every brand should
be running quarterly.

Most brands have never done this. Start by pulling up Shopify billing settings — every app subscription is listed there with its monthly cost. Print the list or copy it to a spreadsheet. That's step one: just knowing what you're paying for.

The pattern that shows up in nearly every audit: 20% of the apps are generating 80% of the attributable value. Those are almost always the core email platform, the attribution tool, and the customer service tool. Everything else exists on a spectrum from clearly worth it to clearly not, with the middle full of apps that were once useful and have quietly become background noise.

One more thing the audit surfaces: apps with free tiers that you've long since outgrown, and are now silently charging you for features you could get cheaper elsewhere. Run a search for "was this on a paid tier when we switched?" for anything you've had installed for more than 18 months. Vendors change pricing frequently. What was free or cheap in 2023 may have moved to a higher tier without anyone noticing.

The underlying principle: your stack should be the smallest set of tools required to run the business effectively at your current revenue tier. Not the most sophisticated. Not the most impressive to show investors. The smallest effective set. Adding apps is easy. Removing them is what separates operators from collectors.