Game Shops Online

Selling a video game online means one of three things, and they have almost nothing in common operationally. The first is a platform store, where a large operator runs the shop, the payment system, the download servers and the customer relationship, and a developer supplies the game. The second is direct sales, where a studio runs its own site and keeps the whole relationship. The third is selling inside a game that has already been acquired, which is now the largest of the three by revenue.

Those differences decide almost everything else. A studio on a platform store has no control over discovery, pricing mechanics or the payment step, but also no chargebacks to fight, no tax registrations to maintain and no fraud team to hire. A studio selling direct controls all of it and carries all of it. Most sell through both, and the split is a strategic decision rather than an accident.

The third model changed the industry more than the other two. When the thing being sold is inside a game that players already own, the acquisition cost is zero, the payment relationship already exists, and the item can be produced in unlimited quantity. That is a fundamentally different economics from shipping a boxed product, and it explains most of what has happened to game business models in the last decade.

On a typical platform sale the operator takes a percentage of the gross price, commonly around thirty percent, though every major platform now has reduced tiers for smaller revenues or for long-running titles. Out of what remains, value added tax or sales tax has usually already been removed, since most platforms quote prices including tax and account for it themselves. What reaches the studio is therefore well under half of what a headline discount suggests.

Then there are the costs that never appear on a statement. Payment processing sits inside the platform fee but not inside a direct sale, where it is typically a small percentage plus a fixed amount per transaction. That fixed component is why very cheap items are disproportionately expensive to sell, and why minimum prices and bundles exist at all.

The argument over the split is really an argument about what the fee buys. It covers hosting, global payment coverage, tax compliance in dozens of jurisdictions, fraud handling, refund processing and, crucially, an audience that is already present. A studio that leaves a platform keeps the percentage and inherits every one of those obligations, which is why most of the studios large enough to leave have chosen not to.

A studio selling from its own site keeps far more per sale, owns the customer relationship, and can run promotions on its own terms. It also has to build or rent a shop, connect a payment processor, handle tax in every jurisdiction it sells into, deliver downloads reliably, and answer support mail. None of that is exotic any more, because hosted commerce platforms handle most of it for a monthly fee, but all of it is work that did not exist before.

The decisive factor is usually traffic. A platform store supplies visitors; a direct site does not. A studio with a mailing list, a community and press coverage can sell direct profitably from the first day. A studio without those has built a shop that nobody walks past, and the higher margin on zero sales is still zero. This is the most common miscalculation in the field.

The usual resolution is to use both deliberately: the platform for reach and for players who want everything in one library, the direct site for the people who follow the studio, for editions the platform will not carry, and for merchandise. Handled that way the direct shop does not compete with the platform, it serves a different audience that arrived by a different route.

The first question is what is this, and it has to be answered in the first screen without scrolling. For a game that means a short line of text and moving footage, because nothing describes a game as efficiently as watching it being played. Static images that could be from any title in the genre are the most common failure, and the most expensive one, because a visitor who cannot tell what the thing is leaves before reaching anything else.

The second is will it run, which is why system requirements sit high on every well built page and why platform and language support are stated rather than implied. The third is what does it cost, including whether the price shown includes tax, because a number that changes at the payment step is the single most reliable way to lose someone who had already decided.

The fourth is can I trust this, and it is answered by everything else on the page: whether the studio is named, whether a returns policy exists and is readable, whether reviews are present and mixed rather than uniformly positive. Uniform praise reads as fake to most people now, and a page with a few critical reviews visible generally performs better than one with none.

Regional pricing exists because purchasing power varies enormously and a single global price would put a game out of reach in much of the world while leaving money on the table in the wealthiest markets. Platforms publish recommended matrices that convert a base price into every currency they support, adjusted for local income rather than by exchange rate alone. A studio that ignores the matrix and converts arithmetically usually finds entire regions simply do not buy.

The obvious weakness is arbitrage: an item priced low in one market being acquired there and used elsewhere. Platforms respond with regional locks, payment-origin checks and periodic price rebalancing, none of which are popular and all of which exist because the alternative is a single high global price. It is a trade-off between access and leakage, and every operator sits somewhere different on it.

Seasonal sales are a separate mechanism and they are more powerful than most first-time operators expect. Deep discounts during a major event can produce more total revenue than months of full-price sales, because the event supplies attention that the studio could not buy. The risk is training an audience to wait, which is why mature studios keep a discount rhythm predictable in timing but restrained in depth.

In the European Union a consumer normally has fourteen days to withdraw from a distance purchase. Digital content is the exception: that right can be waived, but only if the customer explicitly agrees before the download begins and acknowledges losing it. This is why a checkbox appears before an immediate download, and why a shop that skips the checkbox has not extinguished the right at all.

Elsewhere the rules differ. Several jurisdictions have required platforms to offer returns for digital games after enforcement action, which is where the familiar pattern of a time limit plus a play-time limit came from. Others leave it entirely to the operator. The practical result is that the policy a player experiences is usually the platform policy rather than the law, and it is often more generous than the law requires.

For a studio selling direct, the workable posture is a clear written policy, generously applied. Most requests are genuine, the administrative cost of arguing exceeds the value of a single sale, and a refused return that becomes a payment dispute costs far more than the sale was worth. The exception is the pattern of repeated requests from one account, which is a fraud signal rather than dissatisfaction.

Card payments fail for ordinary reasons far more often than people realise: an expired card, a bank declining a cross-border transaction, a strong authentication step abandoned halfway. In many markets a double-digit percentage of first attempts does not complete. Most of those customers intended to pay and simply gave up, which makes recovery one of the highest-value and least glamorous parts of running a shop.

Recovery means small, unexciting things. Show the real reason a payment failed instead of a generic error. Offer a second method immediately rather than sending someone back to the start. Keep the basket intact. Send one reminder, not five. Each of those individually moves a small percentage, and together they usually move more than any redesign of the page itself.

Method coverage matters just as much and is often neglected. Card penetration varies enormously by country, and in several large markets bank transfer schemes or wallets are the default rather than the alternative. A shop that offers only international cards is quietly closed to a large part of the world, which usually shows up as unexplained regional weakness rather than as an obvious error.

Digital goods are attractive to fraud because delivery is instant and irreversible. A stolen card used for a download produces an immediate loss that cannot be recovered by refusing to ship. When the real cardholder disputes the charge, the operator loses the item, refunds the money and usually pays a dispute fee as well, so a single fraudulent sale can cost several times its value.

Defences are layered rather than single. Velocity checks catch one card used repeatedly or one account trying many cards. Mismatches between billing country, network location and account history raise a score rather than a verdict. Strong customer authentication, now mandatory for much of Europe, shifts liability to the bank when it is used properly, which is the single most effective change of recent years.

The failure mode to avoid is over-blocking. A rule tuned to stop all fraud will also stop paying customers, and the lost revenue from false rejections routinely exceeds the fraud it prevented. The mature approach is to score rather than block, to send only the uncertain middle to an extra verification step, and to measure both sides of the trade rather than only the fraud number.

Merchandise looks like an easy extension of a game shop and is not. A physical item has to be manufactured ahead of demand or produced on order, stored, picked, packed, shipped, tracked and sometimes returned. It has weight, a customs classification and a destination that may charge duty on arrival. None of that exists for a download, and a team used to digital delivery usually underestimates all of it.

Print on demand removed most of the risk and most of the margin at the same time. A supplier holds no stock, prints when an order arrives and ships directly, so a studio can offer twenty designs without buying any inventory. The cost is a much thinner margin per item and less control over quality and delivery times, which is a reasonable trade for a small operation and a poor one at scale.

The decision usually rests on why the merchandise exists. If it is a revenue line it needs volume, and volume means committing to inventory. If it is a community gesture, print on demand is exactly right and the margin barely matters. Studios that confuse the two end up with a garage full of shirts in sizes nobody ordered.

A live service title carries its own shop, usually with a rotating selection, a premium currency and a seasonal pass. The currency layer exists for several reasons at once: it decouples the price a player sees from real money, it allows bundle sizes that never quite match item prices, and it lets the operator run a single internal price while real-world prices vary by region and platform. Players notice the leftover balance, and that is understood by everyone involved.

Rotation is the other main mechanism. A shop that shows everything at all times converts poorly, because there is no reason to act today. A shop that shows a small selection for a limited window converts far better, which is why the design is now near universal. It is also the aspect regulators have looked at most closely, alongside randomised contents.

Randomised purchases have attracted real legal attention. Several jurisdictions have restricted or prohibited them, others require published probabilities, and platform rules now generally require disclosure regardless of local law. The industry has moved substantially toward direct purchase of a known item and toward passes with visible tracks, which is a better fit for both the rules and how players say they prefer to spend.

A platform store contains far more titles than any person will ever see, so placement is the real currency. Front page features, seasonal event inclusion and algorithmic recommendations decide most of what sells, and all of them respond to early signals: wishlist additions before release, the rate at which a page converts visitors, and review velocity in the first days. A launch is largely a bet on producing those signals quickly.

Wishlists are the clearest example of a mechanism players use casually and studios watch obsessively. They are a demand estimate the operator can act on, they trigger a notification at release and at discounts, and their growth rate before launch is the single most quoted predictor inside the industry. Almost every marketing beat before release exists to move that number.

Reviews behave differently from other retail. Volume matters more than average score for visibility, recent reviews are weighted more heavily than old ones in most systems, and a title that recovers from a poor launch can genuinely reverse its standing. This is why patch notes and visible responsiveness after release are a commercial activity as much as a technical one.

Small operations that last tend to share a few unglamorous traits. Their fixed costs are low enough that a quiet month is survivable, which usually means hosted infrastructure rather than something built in house. They sell to people who came back, because acquiring a new customer costs several times what keeping one does, and a shop dependent on a constant supply of strangers is dependent on an advertising market it does not control.

They also answer support quickly. In a small operation support is the product experience, and the response time on a payment problem or a delivery failure is what people actually remember. It is the cheapest available differentiator and the one large operators are structurally worst at, which is why it is where a small shop should spend its attention.

Finally, they keep their records straight. Tax registration thresholds, the reconciliation between what the payment processor says and what the shop believes, and a clean separation between money held for customers and money that has been earned: the failures that actually close small shops are usually accounting failures rather than commercial ones. The trade is not glamorous, and neither is surviving in it.