The 2026 stack, compared on landed cost
19 tools across the five layers that bill you per order. Every row shows the number on the invoice, the thing the tool is actually good at, and the fee that bites later.
Tool stacks get evaluated as subscriptions. They should be evaluated as cost per order, because that is how they appear in your contribution margin. A $29 tool that removes 40 seconds of manual work per order is cheap; a $299 platform that adds 1.2% to your checkout rate can be far more expensive.
Each layer below follows the same shape: what you are buying, the question that decides it, then the tools most small and mid-size cross-border stores actually run — with the catch that shows up six months after signing up.
1. Storefront
The platform that hosts the catalogue and the checkout. Subscription cost is the easy part; the app stack it drags in is usually 2–4× the plan price.
What does the whole stack cost per order at 1,000 orders a month?
Basic to Advanced; $1 trial month, then list price
Good atFastest launch, deepest app ecosystem, checkout that converts.
Watch outApp subscriptions usually exceed the plan. Add 0.5–2% for apps that charge on GMV.
Good atCheapest at volume, full data ownership, no platform cut.
Watch outYou own security, updates and performance. Developer time is the real cost.
Good atNo transaction fees on any plan; strong B2B and multi-currency features.
Watch outSmaller theme and app market than Shopify; migrations are painful.
Good atTotal control over speed and UX; no per-order platform fee.
Watch outEngineering cost dominates below roughly $1M GMV. Only worth it with a technical team.
Below ~$100k GMV a year, platform choice changes your margin by a few percent at most. Packaging, returns and paid acquisition move it far more — do not spend a month here.
2. Checkout & payments
Card processing, currency conversion and payout fees. The headline rate is never the rate you pay once cross-border and FX are included.
What is the effective rate on last quarter's mix, including FX and chargebacks?
US cards; +1.5% international, +1% currency conversion
Good atBest developer experience, broadest card coverage, clean reporting.
Watch outCross-border surcharges stack silently. A 2.9% headline is 4.1% on a EU-issued card.
Good atBuyer trust at checkout; meaningful share of US/EU shoppers still prefer it.
Watch outHigher rate and longer settlement. Holding periods on new accounts are common.
Region dependent; avoids Shopify's third-party gateway fee
Good atRemoves the extra 0.5–2% Shopify charges when you use an outside gateway.
Watch outAvailability and rates vary by country; payouts follow the platform's calendar.
Good atCheap multi-currency collection and supplier payouts.
Watch outBanking layer, not a consumer checkout. Useful for settlement, not conversion.
This is the layer where a single percentage point is worth more than every subscription on the page. Model your own effective rate before switching gateways for a 0.1% headline difference.
3. Fulfilment & shipping
Pick-and-pack, storage and carrier rates. Almost always quote-based, which is why comparison pages skip it — and why dimensional weight quietly decides your cost.
What is the all-in cost per order at your box size, including storage and returns?
Plus storage, plus carrier rate
Good atMulti-node network, integrations with most storefronts, transparent dashboards.
Watch outStorage and inbound receiving fees grow with slow-moving SKUs.
Good atUS and EU footprint, strong peak-season handling.
Watch outContract terms and minimums; check the exit clause before you migrate a catalogue.
Good atFull control, no minimums, immediate changes to packaging.
Watch outDimensional weight and fuel surcharges decide the invoice. Check the divisor in your contract.
Good atFast US delivery without owning a warehouse; uses your FBA inventory.
Watch outFee schedule changes annually and packaging is Amazon-branded unless you pay extra.
Before negotiating rates, fix the box. Cutting 2 cm off one dimension often beats a negotiated discount, and it compounds on every parcel.
4. Post-purchase
Returns, exchanges, tracking and support. The cheapest retention lever in ecommerce — and the layer most stores still run through email.
Cost per return processed, and cost per ticket resolved?
Good atExchanges over refunds, which protects revenue instead of reversing it.
Watch outPricing scales with return volume; model it against your return rate, not order count.
Good atCheap entry point, tracking pages that cut 'where is my order' tickets.
Watch outAdvanced automation and branded portals sit on higher tiers.
Good atConsolidates email, chat and social into one queue with order context.
Watch outPer-seat pricing punishes seasonal hiring; check the automation limits.
A returns portal is not a cost centre. Exchanges instead of refunds typically recover 15–30% of would-be refunded revenue — usually more than the tool costs.
5. Measurement
Analytics and attribution. The only layer where spending more is often worse: expensive attribution tools flatter numbers that a return-adjusted recount deletes.
Does reported ROAS survive a return-adjusted recount?
Good atBaseline traffic and funnel data; free forever; consent-mode aware.
Watch outAttribution windows and modelled conversions make channel totals unreliable.
Good atRecovers events lost to browser tracking protection; better consent control.
Watch outSetup and maintenance time; a misconfigured container silently drops data.
Good atBlended MER and channel-level profit in one screen; fast for daily decisions.
Watch outAttribution is a model, not truth. Below ~$50k/month ad spend the subscription rarely pays for itself.
Good atQuery-level demand data — the only place you see what people actually typed.
Watch outNo revenue data; sampling on large sites.
Track contribution margin per order, not blended ROAS. A dashboard that reports gross revenue will keep telling you a losing campaign is profitable.
Three fee traps worth checking before you switch
1. The payment rate is not the rate you pay
A 2.9% headline becomes 4.1% once currency conversion, cross-border surcharges and chargebacks are added. Ask every processor for the effective rate on last quarter's volume, not the published card.
2. Dimensional weight is priced into every 3PL quote
Fulfilment quotes assume a box size. When packaging changes, cost changes — often without a new quote. Check the divisor in the contract and run the numbers in the volumetric weight calculator before you sign.
3. Reported ROAS is not return-adjusted
Platforms count gross revenue at purchase. A 15% return rate takes a reported 3.0x down to roughly 2.55x, which is below the floor for many stores. Model it in the break-even ROAS calculator before you blame the creative.
How this page is maintained
Every row carries the pricing model as published, plus the catch we would want to know
about before signing. Rates are re-checked quarterly; the check date sits at the top of the
page and in the stack.ts data file, so nothing here can silently rot for a
year.
Vendors are listed in the order a store would sensibly evaluate them, not in commission order. Where an affiliate program exists, the link is marked Aff and carries a sponsored attribute. No vendor has paid for placement, and no row has been added or moved because of a commission rate.
Sources checked
Prices are indicative list prices for small and mid-size merchants and exclude negotiated rates, annual discounts and regional variation. Confirm with the vendor before you switch.
How this page earns
Some links on this site are affiliate links. If you buy through them we may earn a commission at no extra cost to you. Rankings and calculator results are never influenced by commission rates.