Ecommerce Platform Cost: The Full Bill

Ecommerce platform cost is the one software bill where the subscription is the small part. A plan at an illustrative $79 a month is $948 a year, which any owner can budget without flinching. The same store doing $300,000 in online sales hands roughly $10,200 to payment processing across that same year. The plan is real, and it is about six percent of what selling online actually costs.

That gap is why pricing an online store the way you price a website goes wrong. Our business website cost verdict prices getting a site built, and our website builder cost verdict prices the subscription that keeps a brochure site online. This verdict prices the part neither of them covers: the percentage of every single sale that leaves the business the moment a customer checks out, and everything that stacks on top of it. Model your own store as you read with the true-cost companion.

Ecommerce platform cost, in one answer

If you want one planning number, use this: the whole platform bill for a small online store commonly runs an illustrative 4 to 5.5 percent of online revenue. On $60,000 a year that is roughly $3,100. On $300,000 it is roughly $14,600. On $1.2 million it is roughly $54,700. Those totals include the subscription, payment processing, the app stack, amortised design, and the domain and mailbox lines.

The subscription inside those totals is small and stubbornly stays small: about $348, $948, and $3,600 a year respectively. That is 11 percent of the bill at the smallest store and under 7 percent at the other two. Every one of those figures is an illustrative band drawn from common patterns in the category, not a quote from any platform.

The practical consequence is that comparing store platforms on plan price is comparing them on the least important line. A plan $20 a month cheaper saves $240 a year. A processing arrangement a quarter of a point better saves $750 a year at $300,000 in sales. The rest of this verdict prices each line in turn, and the true-cost companion keeps your own running total visible as you read.

Why selling online prices differently from a brochure site

A brochure website has a cost that sits still. You pay a plan, a domain, maybe an email mailbox, and the bill next December looks like the bill last December whether the business tripled or halved. That predictability is the whole appeal of the category, and it is why our website builder cost verdict can give a three-year total with a straight face.

Selling online breaks that model in one specific way: a percentage of every sale leaves the business at the moment of the sale. Nothing else in a small-business software stack behaves like this. Accounting software does not take a cut of your revenue. A CRM does not skim the deals it helps close. An ecommerce platform, and more precisely the payments layer beneath it, does exactly that, on every order, forever.

Once a line item is proportional to revenue, it stops being a subscription and starts being a margin decision. Growth does not amortise it. Doubling sales doubles it. That single structural fact reorganises the whole cost analysis: the questions that matter become what percentage you pay, what fixed fee sits under it, and how those two combine at your particular average order value. Everything else on the bill is arithmetic around the edges.

The five layers of an ecommerce platform bill

An honest ecommerce budget has five layers, and most published pricing pages show you exactly one of them.

The first is the platform subscription, the monthly plan that buys the catalogue, cart, checkout, and admin. The second is payment processing, a percentage of each sale plus a fixed fee per transaction. The third is any platform transaction fee, an extra percentage some platforms charge when you bring your own processor rather than using theirs. The fourth is the app and plugin stack, the reviews, email, subscriptions, shipping, tax, and search tools most stores end up needing. The fifth is design: a theme, its customisation, and any ongoing development.

Sitting quietly underneath all five are the domain, the mailboxes, and the small recurring odds and ends. They are the smallest layer and the only one that behaves like a brochure site’s bill.

Layers two and three are the ones that make an online store different, because they are the only ones expressed as percentages. Layers one, four, and five are fixed monthly or annual sums that a growing store amortises across more revenue every year. Layer two never amortises. That asymmetry drives every conclusion in this verdict, and it is the same pattern our true cost of business software verdict finds whenever a vendor prices against a customer’s own volume rather than against a seat count.

What the subscription tiers actually buy

Store platforms ladder their plans much like every other software category, and the rungs are consistent enough to plan around even though the names differ. An entry store tier, commonly an illustrative $30 a month, buys a working shop: product catalogue, cart, checkout, order management, and enough shipping configuration to fulfil. It is genuinely sufficient for a store finding its feet.

The middle rung, commonly around $80 a month, is where most growing stores land. It typically adds staff accounts, better reporting, abandoned-cart recovery, gift cards, and, importantly, a lower platform transaction fee. The advanced rung, from an illustrative $300 a month, buys deeper reporting, more staff seats, advanced shipping rate calculation, and the lowest fee tier the platform offers.

There is a fourth rung above that, sometimes called enterprise, priced by negotiation against your volume. If you are reading a cost article you are almost certainly not there yet.

The useful discipline is the same one we apply to every category: list what the store must do, find the lowest rung that does it, and treat every higher rung as a purchase to justify. The twist unique to ecommerce is that the justification often has nothing to do with features. A higher rung can pay for itself purely through a lower fee percentage, which is the arithmetic in the tier-flip section further down.

Payment processing: the line that dwarfs the plan

Payment processing is two numbers pretending to be one: a percentage of the transaction, plus a fixed fee per transaction. A commonly cited illustrative shape for online card payments is 2.9 percent plus 30 cents, though the real figure varies by processor, card type, region, and your own negotiated arrangement. The mechanics of how that money moves, and who takes which slice, are laid out in our merchant account explainer.

Run that shape against real revenue and the scale becomes obvious. A store doing $300,000 a year across 5,000 orders pays 2.9 percent of $300,000, which is $8,700, plus 5,000 fixed fees of 30 cents, which is $1,500. Total processing: $10,200 a year. The $79 a month plan on the same store is $948. Processing is roughly eleven times the subscription.

This is not a platform markup and it is not avoidable by shopping harder. Accepting card payments has a cost, and every route to accepting them carries a version of it. What is avoidable is paying more of it than your volume warrants, which is what the negotiation section is for. What is not avoidable is the arithmetic: this line grows exactly as fast as the business does.

The fixed fee per transaction, and why small baskets hurt

The percentage half of the fee is proportional, so it feels fair at any basket size. The fixed half is not, and it is where small-basket stores quietly lose money.

Thirty cents on a $200 order is 0.15 percent. Thirty cents on a $25 order is 1.20 percent. The published rate is identical in both cases; the effective rate is eight times further from it in the second. Expressed as a formula, your effective processing rate is the percentage rate plus the fixed fee divided by your average order value. That single line explains more about ecommerce margin than any feature comparison ever will.

Run it across a range of baskets on an illustrative 2.9 percent plus 30 cents. A $25 order costs $1.03 to process, an effective 4.10 percent. A $40 order costs $1.46, an effective 3.65 percent. A $60 order costs $2.04, an effective 3.40 percent. A $100 order costs $3.20, an effective 3.20 percent. A $200 order costs $6.10, an effective 3.05 percent.

The store selling $25 items is paying more than a full percentage point above the store selling $200 items on precisely the same contract. This is why raising average order value through bundles, minimums, or free-shipping thresholds is a cost lever and not only a revenue tactic. Every dollar you add to the basket dilutes the fixed fee.

The platform transaction fee for using an outside processor

Some platforms charge an additional percentage on sales when you use a payment processor other than the one they sell. It sits on top of whatever your processor already takes, so it is straight additional cost for an identical transaction, and it is the single most misunderstood line on a store bill.

The fee commonly steps down as you climb the plan tiers. An illustrative structure runs 2 percent on the entry plan, 1 percent on the middle plan, and 0.5 percent on the advanced plan, waived entirely if you use the platform’s own payments product. Those specific numbers are illustrative; the pattern of a fee that shrinks with tier is the part worth checking on any platform you weigh.

Because the fee is a percentage and the plan gap is a fixed sum, the two cross at a computable revenue. Moving from the entry plan to the middle plan costs an extra $600 a year in subscription and saves 1 percent of sales, so the break-even is $60,000 a year in revenue. Moving from the middle plan to the advanced plan costs an extra $2,652 a year and saves 0.5 percent, so the break-even is roughly $530,000 a year.

Above those thresholds the pricier plan is genuinely the cheaper total. Below them it is not. There is no judgement in that, only arithmetic, and the true-cost companion will run it against your own figures.

Effective cost per order, and how to compute yours

The single most useful number in ecommerce budgeting is not a monthly total. It is the cost of one order. It is the number that tells you whether a promotion is profitable, whether a $19 product is worth listing at all, and whether the platform bill is a rounding error or a problem.

Compute it in two layers. The processing layer is your percentage rate multiplied by average order value, plus the fixed fee. The fixed layer is your monthly subscription plus monthly app spend, divided by your monthly order count. Add the two and you have your all-in platform cost per order.

Worked through on the middle store used throughout this verdict: a $60 basket costs $2.04 to process, and the $79 plan plus $180 of apps across roughly 417 orders a month adds about 62 cents. All-in, about $2.66 an order, which is 4.44 percent of the order. Add amortised design and the domain and it lands near $2.92, or 4.87 percent. Notice that the dollar cost per order rises as baskets grow while the percentage falls, because bigger baskets carry more absolute fee and less relative fee.

How the tier decision flips once fees enter

Compare two store plans on subscription alone and the cheaper one always wins, because a smaller number is a smaller number. Add the fee layer and the ranking can invert completely, which is the most practically valuable thing in this verdict.

Take an illustrative store at $300,000 a year using an outside processor. On the entry plan it pays $348 in subscription and a 2 percent platform fee of $6,000, a total of $6,348. On the middle plan it pays $948 in subscription and a 1 percent fee of $3,000, a total of $3,948. The pricier plan is $2,400 a year cheaper. On the advanced plan it pays $3,600 in subscription and a 0.5 percent fee of $1,500, a total of $5,100, so the middle plan is still the right answer at this revenue.

Push the same store to $1.2 million and the ranking moves again. The middle plan costs $948 plus $12,000, which is $12,948. The advanced plan costs $3,600 plus $6,000, which is $9,600. The advanced plan is now $3,348 a year cheaper, exactly as the $530,000 break-even predicted.

The rule that falls out is simple and absolute: never compare store tiers on plan price. Compare on plan price plus every percentage the tier charges, evaluated at your actual revenue. It is the same discipline our true cost of business software verdict applies to per-seat pricing, moved from headcount to sales.

The app and plugin stack most stores end up needing

Every store platform ships a capable core and a marketplace for everything else, and the marketplace is where the bill quietly grows. The core handles catalogue, cart, checkout, and orders. It does not usually handle product reviews, email marketing and automation, recurring subscriptions, discounted shipping labels, sales tax calculation across jurisdictions, or search and filtering good enough for a large catalogue.

Each of those is a real business need rather than a nice-to-have, and each has a monthly price. An illustrative stack for a store at $300,000 a year might run reviews at $15, email and automation at $45, subscription orders at $30, shipping and labels at $20, tax calculation at $30, and site search and filtering at $40. That is $180 a month, or $2,160 a year.

Compare that to the $948 subscription and the picture inverts: the apps cost more than twice the platform. This is not a scandal, and most of those tools genuinely earn their keep. It is a budgeting failure only when the plan price is treated as the software cost of running a store, which is how nearly every ecommerce budget we see is built.

How per-app monthly fees compound

App spend grows in a specific pattern that is worth recognising because it is so easy to miss. Apps arrive one at a time, each in response to a real problem, each priced at a number too small to require a decision. Nobody approves a $2,160 annual app budget. Six people approve six $15 to $45 monthly line items across eighteen months.

Three mechanics make it compound. The first is usage tiering: many store apps price against order volume, contact count, or catalogue size, so the same app costs more next year without anyone changing anything. The second is overlap: a store commonly ends up with two tools that both send email, or a search app duplicating a feature the higher platform tier already includes. The third is inertia, the same one our SaaS cancellation walkthrough exists to break: apps are installed in busy moments and removed almost never.

The antidote is an annual audit with two questions per app. Would we install this today at its current price, and does anything else we already pay for do this job? On a $180 a month stack, cancelling two overlapping apps at $30 each is $720 a year, which is more than three quarters of the entire platform subscription. That is the kind of saving that is invisible when you budget by plan price.

Theme and design costs, one-off versus custom

Design is the layer with the widest range on the whole bill, because it spans free to five figures with nothing structurally forcing a store toward either end.

At the low end, a platform’s free themes are genuinely usable and cost nothing. Above them, a premium theme is a one-time purchase, commonly an illustrative $150 to $400, that buys a better starting layout and more built-in sections. Above that, paying a freelancer to customise a purchased theme to your brand commonly runs an illustrative $500 to $3,000 depending on how far from the template you go. Full custom design and front-end build starts around $5,000 and has no ceiling, which is the territory our small business website design verdict covers in detail.

The right way to carry design in an ecommerce budget is amortised. A $3,000 theme-plus-customisation project spread across three years is $1,000 a year, which sits it fairly alongside the recurring lines instead of distorting year one. On the middle store in this verdict, that $1,000 is about 7 percent of the annual bill, more than the platform subscription itself.

The trap is treating design as one-and-done. Stores redesign, seasonal campaigns need landing pages, and new product types need new templates. Budget a modest ongoing design line rather than a single launch expense followed by an unpleasant surprise in year two.

Hosted versus self-hosted: the honest comparison

The hosted-versus-self-hosted argument generates more heat than any other question in this category, and most of it is spent on the wrong fifth of the bill.

Start with what does not change. Payment processing is the same on both routes. A self-hosted store at $300,000 a year with 5,000 orders still pays roughly $10,200 to process cards, exactly as the hosted store does. That is about 70 percent of the hosted store’s total bill, and it is entirely outside the argument. Design costs are also broadly route-independent, and so are the domain and mailboxes.

What genuinely differs is a narrow band: the platform subscription and marketplace apps on one side, versus hosting, extension licences, security tooling, and developer time on the other. On the middle store, that band is about $3,100 a year hosted. The next section prices the self-hosted equivalent, and the honest answer is that it is usually higher in cash and lower in cash only when the maintenance work is genuinely free to you.

What self-hosting actually costs to run

Self-hosted store software is typically free to download, which is where the confusion starts. Free software is not a free store, because everything the hosted subscription bundled now arrives as a separate invoice or a separate chore.

An illustrative self-hosted equivalent of the middle store looks like this. Managed hosting capable of running a real store under traffic, an illustrative $150 a month, or $1,800 a year. Premium extension licences for the same six jobs the hosted store’s apps did, an illustrative $1,200 a year in renewals. Security, backups, and uptime monitoring, an illustrative $200 a year. Developer time for updates, breakages, and compatibility work, an illustrative 24 hours a year at $100 an hour, which is $2,400. Total: about $5,600 a year against roughly $3,100 hosted.

That $2,500 gap is not an argument that self-hosting is wrong. It is an argument that self-hosting trades money for control and skill. If you or someone on the team does the maintenance competently, the developer line collapses and self-hosting wins outright. If that work becomes a retainer, it does not.

The line that decides it is rarely on any pricing page: who is going to apply the update at 9pm when an extension breaks checkout. Answer that honestly and the route usually picks itself.

Negotiating your processing rate once volume justifies it

Because processing is the biggest line, it is also the line where negotiation pays best. Most small stores start on flat-rate pricing, which is simple, transparent, and priced for the convenience of never having to think about it. That trade is correct early and gets expensive later.

The arithmetic of why it matters: on an illustrative $1.2 million a year across 15,000 orders, shaving 0.2 points off the percentage is $2,400 a year, and taking 5 cents off the fixed fee is another $750. That is $3,150 for a few conversations, and it recurs every year. On the $300,000 store the same moves are worth $600 and $250, smaller but still comparable to the entire annual subscription.

Below roughly $250,000 a year in card sales there is usually little room, so the honest advice is to revisit this as you grow rather than fight for it early. When you do ask, bring twelve months of statements, ask specifically about interchange-plus style pricing rather than a lower flat rate, and get a competing quote so the request has weight behind it. The general approach in our SaaS negotiation verdict transfers cleanly, with one difference: here the vendor can see your volume growing, which makes the timing of the ask your strongest card.

Chargebacks, refunds, and the fees that do not come back

Two quieter fee behaviours deserve a place in the budget because they surprise owners at exactly the wrong moment.

The first is refunds. When you refund a customer, the money goes back to them, but the processing fee on the original sale commonly does not come back to you. On an illustrative $60 order that is $2.04 gone on a sale that generated no revenue. A store with a 5 percent refund rate on 5,000 orders is absorbing roughly $510 a year in fees for transactions that were reversed. Policies differ by processor, so confirm yours rather than assuming either way.

The second is chargebacks, where a customer disputes a charge with their card issuer. Beyond losing the sale and often the goods, a chargeback commonly carries an administrative fee per case, and a persistently high dispute rate can trigger higher rates, reserves, or account review. The mechanism and the risk posture around it are covered in our merchant account explainer.

Neither line is large enough to change a platform decision, and neither is worth panicking about. Both are worth a small budgeted allowance rather than a quarterly surprise, and both reward the boring operational fixes: clear product photography, honest descriptions, recognisable billing descriptors, and fast responses to customer emails before a dispute is filed.

Cross-border, currency, and the fee add-ons

If you sell beyond your home market, two further percentages can attach to the same transactions, and they are easy to miss because they appear as separate line items rather than as a higher headline rate.

The first is a cross-border or international card fee, commonly an illustrative additional 1 percent or so when the customer’s card was issued in a different country from your account. The second is a currency conversion fee, commonly an illustrative 1 to 2 percent, applied when you present prices in a customer’s currency and settle in your own. They stack: a store where a quarter of orders come from abroad might see its blended effective rate rise by roughly half a point overall even though its domestic rate never changed.

Multi-currency presentation is also often a paid feature or a higher tier on the platform side, so the same decision touches two layers of the bill at once.

None of this argues against selling internationally, which usually adds far more revenue than fee. It argues for measuring your blended effective rate rather than quoting the domestic one, because the blended figure is the one that actually appears in your margin. Pull three months of settlement reports, divide total fees by total sales, and use that number in every calculation from then on.

The true annual cost at three revenue levels

Here is the whole model assembled at three revenue levels, with every line illustrative and every figure derived from the ones above.

The small store does $60,000 a year across 1,200 orders at a $50 average basket. Subscription $348, processing $2,100, apps $480, amortised theme $60, domain and mailbox $90. Total $3,078, which is 5.13 percent of revenue and $2.57 an order. The subscription is 11 percent of the bill.

The middle store does $300,000 across 5,000 orders at a $60 basket. Subscription $948, processing $10,200, apps $2,160, amortised design $1,000, domain and mailboxes $300. Total $14,608, which is 4.87 percent of revenue and $2.92 an order. The subscription is 6 percent of the bill.

The large store does $1.2 million across 15,000 orders at an $80 basket, with a negotiated 2.5 percent plus 25 cents. Subscription $3,600, processing $33,750, apps $7,200, design and development $9,000, domain and mailboxes $1,200. Total $54,750, which is 4.56 percent of revenue and $3.65 an order. The subscription is 7 percent of the bill.

Three stores, twenty times the revenue between the smallest and largest, and the subscription never once reaches an eighth of the bill.

What the bill looks like as a share of revenue

Shares tell the story more clearly than totals, because they are what actually survives contact with a growing business.

The revenue share drifts down slowly as a store grows, from 5.13 percent to 4.87 percent to 4.56 percent across the three examples, because the fixed layers spread across more sales while processing stays proportional. That drift is real but modest, which is why the 4 to 5.5 percent planning band holds across a wide revenue range.

Whether that share is comfortable depends entirely on gross margin. At a 50 percent margin, a bill of 4.87 percent of revenue is consuming about 9.7 percent of gross profit before a single other operating cost. That is the number to put in front of anyone who thinks the store platform question is a $79 decision.

The costs that sit outside the platform bill

For completeness, several genuine costs of selling online sit outside everything priced above, and leaving them unlabelled is how a store budget ends up wrong by a multiple rather than a margin.

Shipping and packaging are usually the largest of them, and they behave like processing in that they scale with orders rather than with time. Returns handling carries both a shipping cost and a restocking labour cost. Inventory itself is working capital rather than an expense, but it competes for the same cash, which is why the tooling in our inventory management software verdict tends to earn its price quickly. Marketing, whether paid acquisition or email, is typically the largest discretionary line of all.

There is also the in-person side. Stores that sell at markets or in a shop as well as online carry hardware, a separate in-person processing rate, and often a second software subscription, which our POS system cost verdict prices properly.

The reason to name these rather than bundle them is that they respond to entirely different levers. You cannot negotiate your way out of packaging, and you cannot pack your way out of a bad processing rate. Keep them in separate buckets and each one stays fixable.

How to audit an ecommerce platform bill you already pay

If you already run a store, the highest-return hour available to you is an audit, because you are working from real figures rather than planning bands.

Start by pulling twelve months of processor settlement reports and dividing total fees by total sales. That single ratio is your blended effective rate, and it is almost always higher than the rate you believe you are on, because refunds, cross-border cards, and currency conversion are all in it. Then divide total fees by total order count to get your real cost per order.

Next, list every recurring charge against the store: platform plan, every app, hosting if self-hosted, and any retainer. Put the annual figure beside each. Sort descending. Most owners find that their sorted list looks nothing like their mental model of it, and that two or three apps they had forgotten are larger than the platform plan.

Finally, run the tier arithmetic. Multiply your annual sales by the platform transaction fee on your current tier and on the next one up, add each tier’s subscription, and see which total is smaller. Then decide whether your volume now justifies a processing conversation. Feed your own figures into the true-cost companion as you go and the four steps take about an hour between them.

The mistakes that inflate an ecommerce platform bill

Five patterns account for most of the difference between a lean store bill and a bloated one, and all five are ordinary rather than foolish.

Comparing platforms on plan price is the first and largest. It optimises the smallest line while ignoring the percentages, and it routinely leads stores onto a cheap tier whose transaction fee costs several times the saving.

Staying on a tier past its break-even is the second. The break-even moves as you grow, so a decision that was correct at $50,000 in sales can be costing money at $400,000 without anything having changed except success.

Letting the app stack accumulate is the third, and it is the easiest to fix. An annual audit with a genuine willingness to cancel typically finds 10 to 20 percent of the app line.

Ignoring average order value as a cost lever is the fourth. Raising a $40 basket to $60 does not only add revenue, it cuts the effective processing rate from 3.65 to 3.40 percent.

Assuming free software means a free store is the fifth. Self-hosted platforms are a legitimate route with real advantages, but the invoice moves rather than disappears, and the developer hours are the part that gets forgotten.

A worked example: two stores at the same revenue

Two stores, identical $300,000 in annual sales, identical $79 plan, identical $180 a month of apps. The only difference is basket size, and the bills are not the same.

The first sells small consumables at a $30 average order, so it processes 10,000 orders a year. At an illustrative 2.9 percent plus 30 cents, each order costs $1.17 to process, an effective 3.90 percent, and the year’s processing comes to $11,700. Adding the subscription and apps gives a recurring platform bill of $14,808, or 4.94 percent of sales, at $1.48 an order.

The second sells higher-value goods at a $150 average order, so it processes 2,000 orders. Each costs $4.65, an effective 3.10 percent, and the year’s processing is $9,300. Its recurring platform bill is $12,408, or 4.14 percent of sales, at $6.20 an order.

Same revenue, same software, $2,400 a year apart. The entire difference is 8,000 extra fixed fees of 30 cents each. The small-basket store is not doing anything wrong, but it should know that its checkout costs it a full 0.8 points more of revenue, and that a bundle or a free-shipping threshold pushing the average basket to $45 would recover a meaningful share of that. Run your own basket size through the true-cost companion and watch the effective rate move.

The bottom line

Ecommerce platform cost is not a subscription question dressed up with extras. It is a percentage-of-revenue question with a subscription attached. The plan, at an illustrative $30, $80, or $300 a month by tier, is the line every comparison article ranks and the line that matters least: across three stores from $60,000 to $1.2 million in sales it never exceeded 11 percent of the real bill, and twice sat near 6 percent.

The bill that matters is roughly 4 to 5.5 percent of online revenue, and payment processing is about 70 percent of it. That means the levers worth your attention are the percentage rate, the fixed per-transaction fee, the platform transaction fee your tier carries, your average order value, and an app stack that is audited rather than accumulated. Get those five right and the plan price becomes what it should have been all along: the smallest decision on the page.

This verdict from VetLoft is written to help you budget an online store, not to price any named platform, processor, or app. Plan tiers, processing rates, fixed per-transaction fees, platform transaction fees, and app pricing all vary by provider, region, card mix, and negotiated arrangement, and they change without notice, so every dollar and percentage above is an illustrative model built to be internally consistent rather than a quote you can hold anyone to. Before you commit, pull the current pricing and fee schedule for each platform and processor you are weighing, confirm the transaction fee at your intended tier in writing, and check your own settlement reports for the blended rate you are genuinely paying.