Glossary of one-tap commerce terms

Ordered by how the terms fit together rather than alphabetically, so reading top to bottom builds the argument the rest of the site makes. Each definition stands on its own, so any single one can be quoted without the rest. Where a term is our coinage rather than established usage, the definition says so — a glossary that quietly promotes its author's vocabulary into the industry's is a glossary you should not trust.

One-tap commerce: One-tap commerce is Forage's term for buying something from a physical surface — a can, a carton, a shelf-edge, a poster — in a single confirmed gesture, with no app install and no typed address. It is our coinage rather than established industry usage, and we use it because each alternative names only a component: "QR commerce" names the code, "connected packaging" names the surface, "one-tap checkout" names only the payment. The thing worth having a word for is the sequence: a printed code resolves to a mobile storefront, a digital wallet supplies the card and the address, and the purchase attaches to a customer record so the second one is easier than the first.
One-tap checkout: One-tap checkout is a checkout that completes in one confirmed gesture, because a digital wallet already holds the card, the billing address, the delivery address and the identity check. The shopper's contribution is a fingerprint, a glance or a passcode. The form that used to sit between them and the order is not filled in faster — it is not shown at all. Apple Pay, Google Pay and Shop Pay are the usual implementations. What matters commercially is that the saving is not the seconds of typing but the abandonment the typing causes, on a phone, one-handed, in a shop or on a bus.
Express checkout: Express checkout is the industry's name for the wallet buttons that sit above or instead of a standard checkout form: Apple Pay, Google Pay, PayPal, Shop Pay. It is broader and vaguer than one-tap checkout, because an express checkout button can still open a multi-step flow, and several do. When a vendor quotes an express-checkout conversion figure, the question the label does not answer is how many taps the flow actually takes for a first-time customer on a phone.
Digital wallet: A digital wallet holds a customer's payment credentials and addresses on their behalf — in the phone or the browser, as Apple Pay and Google Pay do, or in an account, as PayPal and Shop Pay do — and releases them to a merchant after an identity check, as a token rather than a card number. Worldpay's Global Payments Report 2026 puts digital wallets at 56% of global online spending in 2025. For a brand printing a code on a pack, the wallet is why the whole sequence works: it is the only thing that can hand a delivery address to a stranger's checkout without anybody typing one.
App Clip: An App Clip is a small, signed slice of an iOS app that Apple launches without an App Store install, invoked by a code, an NFC tag, a link or a banner. It is a build target inside a full iOS app rather than a product of its own, so shipping one requires an app to put it in; it does not sit on the home screen, and iOS removes it after a period of inactivity. Apple's size ceiling depends on how it is invoked: a code on packaging is a physical invocation, which caps the build at 15 MB for iOS 16 and later, or 10 MB if earlier versions are supported. The 100 MB tier that gets quoted applies only to App Clips invoked digitally.
App Clip Code: An App Clip Code is Apple's own circular scannable marker: a ring of coloured dots around a central icon, encoding a URL, which opens an App Clip from the iPhone camera or from a tap when it is paired with an NFC tag. It is a designed artefact with its own colour, contrast and minimum-size rules, which makes it an artwork sign-off item rather than something generated at the end of a project. An ordinary QR code invokes an App Clip on iOS just as well, so an App Clip Code is chosen for recognisability rather than for capability.
Google Play Instant: Google Play Instant was Android's install-free app runtime, the closest thing to Apple's App Clips, and it has been withdrawn. Google's own documentation states that from December 2025 instant apps can no longer be published through Play, the Play services Instant APIs stopped working, and users are no longer served instant apps by any mechanism; the recommended alternative is a deep link into the installed app. For anyone printing a code, the consequence is that there is now no install-free native path on Android at all, so the code has to resolve to a web page good enough to buy from.
The install tax: The install tax is Forage's term for what you charge a customer when the shortest route to buying from you runs through an app store: find the app, agree to the download, wait for it, open it, make an account, then buy. It is our coinage rather than a standard term. We call it a tax because it is paid before any value is received, and because it is paid by everybody — including the people who abandon partway through and are therefore never counted in the app's own numbers. The argument the phrase exists to make is that a customer should never have to pay it in order to give you money.
Packaging as inventory: Packaging as inventory is Forage's term for treating printed packs as media you have already bought and already shipped: every carton in a warehouse and every can on a shelf is an addressable surface that reaches a customer's hands with no further media spend attached to it. It is our coinage rather than established usage. The point of the reframing is budgetary. A code on a pack competes with paid social for attention but not for money, so the useful question stops being "what will this cost" and becomes "what share of the packs we are printing anyway can carry something worth scanning".
Connected packaging: Connected packaging is the trade term for packaging that carries a scannable trigger — a QR code, an App Clip Code or an NFC tag — so that the pack becomes an addressable, measurable channel instead of a printed surface. It is distinct from smart packaging, which means sensors, freshness indicators and RFID reporting on the condition of the product inside. Connected packaging is about reaching the person holding the pack; smart packaging is about the pack reporting on itself. The two are used interchangeably in trade press and they describe different products.
Scan-to-buy: Scan-to-buy is the pattern where a shopper points a phone camera at a code on a physical surface and completes a purchase in the session that opens, rather than being delivered to a website to start browsing. What separates it from an ordinary QR code on a pack is the destination: scan-to-buy lands on one specific product with a wallet button attached, not on a homepage or a campaign microsite. Almost all of the difficulty sits in that choice, because a code that resolves to a general site inherits every step of a normal mobile checkout.
Guest checkout: Guest checkout is buying without creating an account. It is worth defining carefully, because it is routinely mistaken for the cure for checkout abandonment. Among the people in Baymard Institute's study who abandoned during checkout, 18% cite a required account among their reasons and 40% cite extra costs, so the account is one factor and not the largest. The friction that dominates is the typing rather than the sign-up. A guest checkout with an eleven-field form is still an eleven-field form, and a wallet payment removes the fields whether an account exists or not.
Cart abandonment: Cart abandonment is the share of started baskets that are never paid for. Baymard Institute's documented average, taken across 50 separate studies, is 70.22%. Two cautions travel with that number. It is an average of other people's studies rather than a measurement of your shop, so it is a benchmark and nothing more. And the reasons Baymard records — extra costs, slow delivery, distrust, a required account, a checkout that is too long — are shares of the people who abandoned during checkout, not shares of everyone who intended to buy, so quoting them the other way round overstates every cause.
Loyalty enrolment friction: Loyalty enrolment friction is everything a customer has to do between deciding that a rewards scheme is worth having and actually being in it: find the sign-up, hand over an email address, confirm it, choose a password, install something, remember a card. It is why enrolment figures and participation figures diverge so far, and why schemes end up reported on members rather than on repeat purchases. Removing it means enrolling somebody on the purchase they were already making, rather than asking them to complete a separate task afterwards.
Incrementality: Incrementality is the share of results that would not have happened without the thing being measured: the orders a campaign caused, as opposed to the orders it happened to be standing near. It is the honest counterpart to attribution, which assigns credit for whatever follows a touchpoint whether or not the touchpoint changed the outcome. For a code on a pack the gap between the two is unusually wide, because the people most likely to scan are the people most likely to reorder anyway.
Holdout test: A holdout test measures incrementality by deliberately withholding the thing from a matched group and comparing the two: some regions, stores, batches or customers get the code and an equivalent set does not, and the difference between them is the effect. It is the only design that answers "did this cause anything" rather than "what happened afterwards". For printed packaging it has to be planned before the print run, because once every pack carries a code there is no comparable group left to hold out.
Cannibalisation: Cannibalisation is a new channel taking sales an existing one would have made anyway, which flatters the new channel twice: once in its own reported numbers, and again by quietly reducing somebody else's. A scan-to-buy purchase by a customer who reorders every month is a channel switch, not a new order. For a loyal customer base this is the default outcome rather than an unusual one, which is why the figure worth agreeing before a launch is incremental revenue and not revenue attributed to scans.
Universal link: A universal link is an ordinary HTTPS URL that iOS opens inside an app or an App Clip rather than in the browser, in cases where the domain has vouched for it. The vouching is a JSON file — the apple-app-site-association file — served from the domain's /.well-known/ path, listing which paths belong to which app. It matters for printed codes because it is what lets one URL work everywhere: iOS can claim it and open natively, and every other device treats the same string as a normal link to a web page.
QR code quiet zone: The quiet zone is the empty margin a QR code needs on all four sides so that a camera can tell where the code starts and stops. Crowding it is the commonest reason a code that scanned on a designer's screen fails on a shelf: a fold, a seam, a varnish edge, a nutritional panel or a line of legal text pushed up against the pattern will each do it. The margin is part of the code rather than a design allowance around it, and it only survives artwork review if somebody has said so out loud.
QR code module size: A module is one of the small squares a QR code is built from, and the printed size of a module sets the distance and the conditions under which a camera can resolve the code. Module size is what a printed dimension actually controls, which is why shortening the URL often buys more than enlarging the code: fewer characters need fewer modules, so each module is larger at the same physical width. On packaging, where roughly 20mm across is a sensible floor and curved or coarsely printed surfaces need more, this is what makes the destination URL a print decision rather than only a marketing one.

Where these are argued rather than defined

A definition is not an argument. Each of these terms is worked through properly somewhere on the blog, with the sources attached:

One-tap checkout is already the default — the install tax, express checkout, guest checkout, digital wallets.
App Clips, explained for Shopify brands — App Clips, App Clip Codes, universal links, and the end of Google Play Instant.
Your packaging is already a sales channel — packaging as inventory, connected packaging, scan-to-buy, quiet zones and module size.
The mobile conversion gap — cart abandonment and where the missing orders actually go.
Loyalty without discounting — enrolment friction, and why membership is the wrong measure.
How to prove one-tap actually worked — incrementality, holdout tests and cannibalisation.

For how the pieces fit together in one system, rather than one term at a time, see how Forage fits your stack.

Sources

Four definitions above quote a number. Each one is named in the definition itself and traced here, so a definition can be checked without leaving the page. Superscripts are deliberately absent from the definitions: each one is quoted verbatim in this page's structured data, and a citation marker inside the text would put a footnote number into the quotation.

Baymard Institute, Cart abandonment rate — the 70.22% documented average across 50 studies, and the reason weights among shoppers who abandoned during checkout. Backs cart abandonment and guest checkout. Primary
Worldpay / Global Payments, Global Payments Report 2026 — digital wallets at 56% of global online spending in 2025. Backs digital wallet. Primary
Apple, Maximum build file sizes — the App Clip size tiers, and the digital-invocations-only condition on the largest one. Backs App Clip. Primary
Google, Google Play Instant — deprecation notice: from December 2025 instant apps cannot be published and the Instant APIs no longer work. Backs Google Play Instant. Primary