On this page
concept

Frictionless Flow

Created 2026-10-23 32 connections

Frictionless Flow

In 3D Secure 2 (3DS2) authentication, the frictionless flow is the pathway where the issuer's ACS (Access Control Server) authenticates a transaction silently in the background — without presenting any visible challenge screen to the cardholder. The ACS uses passive risk signals (device fingerprint, transaction history, 150+ data elements) to decide in milliseconds whether authentication can complete without cardholder interaction. It is the dominant path in mature SCA markets and the primary lever for preserving conversion in markets where Strong Customer Authentication (SCA / PSD2) is mandated.

How it works

When a card-not-present transaction is submitted, the 3DS protocol routes it via the card scheme's Directory Server to the issuer's ACS. The ACS conducts a real-time risk assessment across over 150 data elements — device fingerprint, IP geolocation, browser details, transaction history, behavioural signals — and returns one of three outcomes: frictionless (approved without cardholder input), challenge (additional verification required), or failed authentication. (Gr4vy, 2026; Netcetera, 2025-07-22)

A key preparatory step is the 3DS Method (Method URL): a background browser call that collects additional device data before the authentication request is sent. Including this step materially increases the probability of a frictionless outcome by enriching the ACS data set. (Learn Payments, 2023-12-31)

The ACS's decision depends on the quality and completeness of data submitted. Gr4vy (2026) identifies the key data categories: device data, behavioural and account history, order and shipping data, merchant data (MCC, acquirer BIN, requestor name), and transaction data (amount, currency, frequency).

If the bank has enough data to trust the cardholder, authentication completes without further customer input, and liability shift still applies to frictionless transactions (Stripe, 2026 copyright).

Data Only / Data Share Only (DSO)

A variant of the frictionless flow called Data Only (also referred to as Data Share Only or DSO under the EMVCo 3DS specification) submits a 3DS2 request without requesting authentication. Available only for Visa and Mastercard (Adyen, 2026), it uses 3DS infrastructure to share transaction and cardholder data with issuing banks without presenting any challenge, and the issuer uses the additional signal to make better authorisation decisions.

Stripe's AI model determines when a Data Only flow is likely to improve authorisation outcomes and applies it automatically for eligible transactions — no action is required from the merchant integration. (Stripe Changelog, 2026-09-30)

In a Data Only flow, liability does not generally shift to the issuer, because the cardholder has not been presented an authentication step. (Stripe Docs, 2026-09-30)

As of 2026-09-30, Stripe added data_share_only as a new enum value for three_d_secure.result on Charge, SetupAttempt, PaymentRecord, and PaymentAttemptRecord objects. Prior to API version 2026-09-30.endive, DSO authentications were indistinguishable from charges with no 3DS because three_d_secure was null in both cases. (Stripe Changelog, 2026-09-30)

Visa's Digital Commerce Authentication Program (DCAP) is a framework for card-not-present transactions that sends additional cardholder data during authentication to the issuer. For IC+ merchants, DCAP can provide interchange benefits and reduce costs, with specific interchange programs in the US and Canada when combined with network tokens. Required data fields vary by region: cardholder IP address, device ID (except in Europe), billing address, email address, and phone number (except in the US). (Stripe Docs, 2026-09-30)

Frictionless rates — global benchmarks

The 64% global average figure is from Ravelin/Stripe 2024 data. Ravelin's March 2026 report shows rates falling in 76% of countries — the current picture is likely lower for many merchant segments.

A global average frictionless rate of approximately 64% across 37 countries is cited by PAAY sourcing Ravelin and Stripe (2024 data; as-of 2024). Checkout Experts (2026-10-02) cites industry benchmarks commonly at 80–85% frictionless rates for merchants sending complete data. PaymentBrief (2026-04-12, updated 2026-08-31) states frictionless rates of 85–95% are achievable for low-risk transactions at merchants with strong 3DS2 data quality.

Frictionless rate baseline — wide range. PAAY citing Ravelin/Stripe (2024) gives ~64% global average. Checkout Experts (Oct 2026) treats 80–85% as a reasonable practical benchmark. PaymentBrief (April 2026) says 85–95% is achievable for high data-quality merchants. These likely refer to different populations (all merchants vs. optimised merchants vs. low-risk merchants only) but are frequently quoted without that distinction. [paay.co] VS [checkoutexpert.com, 2026-10-02] VS [paymentbrief.com, 2026-04-12]

Frictionless trend direction — the "frictionless paradox". Older sources describe frictionless rates as rising with 3DS2 adoption. Ravelin's Global Payments Report (March 2026) and Quantum Payments (September 2026) report the opposite: frictionless rates fell or plateaued in 28 of 37 countries (76%) despite overall 3DS success rates rising. Quantum Payments attributes this to merchants adding more transaction categories in higher-risk bands; Ravelin attributes it partly to merchants not submitting enough 3DS data elements and issuers tightening risk assessments. The 2026 sources prevail for current state. [ravelin.com/blog/3d-secure-rates-2026] VS [quantumpayments.io/post/the-frictionless-paradox-3ds-success-is-rising-smooth-checkouts-aren-t]

Ravelin's 2026 report shows average 3DS success rates by country (as-of 2026-03-19): UK 95%, Italy 93%, Netherlands 92%, France 91%, Ireland 91%, Czech Republic 91%, Austria 88%, US 88%, Germany 87%, Australia 87%, Poland 87%, Canada 82%, Spain 81%, Belgium 80%, Sweden 76%, Denmark 75%, Brazil 36%.

Ravelin country figures — metric ambiguity. Ravelin labels these as "3DS success rate," which conflates frictionless success and challenge success under the same formula. They are not frictionless-only rates. Secondary sources (Purse, Quantum Payments) quote these in a frictionless-rate context without noting the conflation. [ravelin.com/blog/3d-secure-rates-2026]

Stripe Sessions 2026 data (Amandeep Batra + Cip Blujdea, Stripe): Japan (post-April 2025 mandate) sees ~60% of 3DS requests as frictionless, checkout conversion at ~93%, and dispute rates down >50%. France (mandated) sees ~80% of challenges as biometric, only 6% OTP. US (optional) sees >50% of challenges via SMS OTP, only 24% biometric. (Stripe Sessions 2026, stripe.com/sessions/2026/the-truth-about-3ds)

Ravelin (2026-01) reports 78% of merchants now use SCA Exemptions (up year-on-year), but issuers are increasingly rejecting exemption requests. "Nobody is leveraging exemptions as effectively as they could be."

Conversion impact

PaymentBrief (2026-04-12) cites Stripe analysis suggesting challenge flows reduce checkout conversion by 3–8% relative to frictionless transactions. A well-implemented 3DS2 integration sees a 3–4% conversion impact from challenges; a poorly optimised integration in challenge-heavy markets can lose 8–12% (as-of 2026-04-12).

Gr4vy (2026) states 3DS1's typical conversion impact was 5–15%, while well-implemented 3DS2 keeps conversion impact under 1–2% on affected volume, with a gap of 5–15 percentage points between well- and poorly-implemented deployments.

Adyen's Authenticate upgrade early results (cited by Quantum Payments, 2026-09-24): Just Eat Takeaway.com achieved a 12% increase in payment conversion on eligible traffic; Too Good To Go achieved an 11% relative increase.

US 3DS paradox. Stripe Sessions 2026 reports that in a 2024 US experiment, adding 3DS to some transactions lowered checkout conversion from 87% to 82% even when most transactions would pass frictionlessly — because US issuers may treat the 3DS request itself as a fraud signal. This contradicts the generic framing that frictionless flow is categorically conversion-neutral. The effect appears US-specific and is not observed in mandated European markets. [stripe.com/sessions/2026/the-truth-about-3ds]

Stripe Sessions 2026: "Authorization Boost" (3DS retries) yields ~90 basis points conversion uplift. Data Only in Europe (orders >€250) shows +3.2% conversion vs. no data sharing. (Stripe Sessions 2026)

Merchant levers to improve frictionless rates

Checkout Experts (2026-10-02) identifies actionable steps:

  • Complete data submission. Anecdote (via Entersekt): a merchant with ~80% of 3DS traffic being challenged moved ~80% of those transactions to frictionless after improving data quality alone. PAAY cites Visa's data field mandate as yielding a "+57% frictionless rate lift" when required fields are provided (as-of 2024 data; treat as directional).
  • Wallet tokens. Enabling Apple Pay and Google Pay makes device-tokenised wallet transactions frictionless by construction (though Google Pay cards saved to an account may still require 3DS).
  • Native SDKs. Avoid in-app embedded webviews, which can strip data fields. Use native SDKs or Custom Tabs / SFSafariViewController instead.
  • Per-transaction challenge indicator. Set the 3DS Requestor Challenge Indicator (threeDSRequestorChallengeInd values 05–09 in EMVCo 3DS 2.2+; values 10–11 added in 2.3.1) per transaction rather than a single static preference.
  • Soft decline retry. Retry Mastercard 65 / Visa 1A Soft Decline|soft declines with 3DS rather than showing a failure screen — these are recoverable authentication events. (tomcn.uk, 2026-09-22)
  • Orchestration-based "agnostic" 3DS. Allows authentication to run before processor selection; if a fallback PSP is needed, the result can be reused so the buyer is not challenged again. (Purse, 2026-05-12)

PaymentBrief (2026-04-12): merchants actively managing PSD2 SCA exemptions reportedly achieve frictionless rates 15–20 percentage points higher than those passively processing transactions (no named source for this figure; treat as directional).

Adyen states its model is trained on "trillions" of interactions; for a retail merchant on its platform there is a >90% chance it has seen the shopper before, supporting frictionless outcomes. (Quantum Payments citing Adyen, 2026-09-24)

Challenge completion rates vary by market maturity: 75–85% in European markets with mature banking apps; 55–65% in markets with high SMS reliance and older infrastructure. (Checkout Experts, 2026-10-02)

Native in-app authentication (biometric or push) outperforms SMS OTP on completion; enabling app-based challenges plus smoother web redirects can recover 5–10 percentage points of challenge completion rate. (PaymentBrief, 2026-04-12)

SCA exemptions and frictionless interaction

SCA Exemptions signal a request to skip challenge inside 3DS via threeDSRequestorChallengeInd. Values 01–04 date from 3DS 2.1; values 05–09 require 2.2+; values 10–11 (low value and secure corporate) arrived in 2.3.1. A 2.1 flow cannot signal a Transaction Risk Analysis (TRA)|TRA request inside 3DS. (tomcn.uk, 2026-09-22)

Many PSPs send exemptions in the authorisation message instead (skipping 3DS entirely); when the issuer rejects this, a Soft Decline (Visa 1A or Mastercard 65) is returned. Treating these as a recoverable payment state and retrying with 3DS is recommended. (tomcn.uk, 2026-09-22)

TRA thresholds (EU): acquirer fraud rate below 0.13% → exemption up to €100; below 0.06% → up to €250; below 0.01% → up to €500. Per-PSP on a rolling 90-day window. TRA thresholds (UK): below 0.13% → up to £85; below 0.06% → up to £220; below 0.01% → up to £440. (tomcn.uk, 2026-09-22)

Low-value exemption: EU €30 per transaction, cumulative cap €100 or 5 consecutive transactions; UK £25 per transaction, cumulative cap £85 or 5 consecutive transactions. (tomcn.uk, 2026-09-22)

Liability rules when exemptions are used:

ScenarioLiable party
Frictionless with no exemption appliedIssuer (generally)
Exemption appliedMerchant (generally)
Successful challengeIssuer (generally)
No 3DSMerchant

(Checkout Experts, 2026-10-02)

Forter's 2026 PSD2 report (June 2026 network data): authentication failures on transactions already approved as genuine broke down as: exemption rejection 27.8%, issuer rejection 15.6%, user abandonment 12.6%, technical failure 11.5% — making exemption rejection the single largest cause, more than twice user abandonment. (cited by Checkout Experts, 2026-10-02)

From 12 January 2026, French OSMP rules set the velocity limit for online payments without 3DS at €0.01 for customer-initiated transactions in the EEA, effectively ending "Direct to Authorization" for small amounts. (Purse, 2026-05-12; note: France-specific enforcement of PSD2 RTS)

Regulatory and protocol context

EMVCo released EMV 3DS v2.3 (and 2.3.1) with key additions: more data elements (recurring transaction data, EMV Payment Token data), automated out-of-band transitions, device binding (a recognised device can be trusted for future frictionless decisions), Split-SDK for non-traditional channels (IoT, smart speakers), UI enhancements, and support for WebAuthn and Secure Payment Confirmation (SPC). (EMVCo, 2021-11-12, updated 2024-09-26)

As of 2026, EMV 3DS versions 2.2 and 2.3.x are current; v2.3 was released in 2023 and v2.3.1 in 2024. No EMV 3DS version 2.4 has been confirmed or publicly announced. (Gr4vy, 2026; EMVCo, 2024-09-26)

PSD3 and the Payment Services Regulation agreed texts were published April 23, 2026; the European Parliament's ECON committee approved both in May 2026; application is expected across 2026 and 2027. (Checkout Experts, 2026-10-02)

Visa is raising its non-authenticated transaction fee to 7.5 bps; Mastercard plans a similar increase — scheme incentive shifts are pushing merchants toward 3DS adoption. (Stripe Sessions 2026)

Stripe added a Standalone 3DS API (ThreeDSecure.Authentication resource with create, submit, retrieve, list, and cancel endpoints) that lets merchants authenticate a card using 3DS independently of a Stripe payment, enabling integration of Stripe's 3DS authentication into externally-managed payment flows. (Stripe Changelog, 2026-09-30)

Stripe exposed the network-returned Electronic Commerce Indicator (ECI) at payment_method_details.card.electronic_commerce_indicator on the Charge object for all businesses (supported Visa and Mastercard authorisations). This network-final ECI is distinct from the 3DS ECI at payment_method_details.card.three_d_secure.electronic_commerce_indicator (which represents the ECI used in the authorisation request). Accessible from June 10, 2026; no backfill of earlier transactions. (Stripe Changelog, 2026-09-30)

Key terms

TermMeaning
ACSAccess Control Server — issuer-side component that scores risk and decides frictionless vs. challenge
3DS MethodBackground browser call (Method URL step) that collects device data before the authentication request
DSO / Data OnlyData Share Only — shares 3DS data with issuer without requesting authentication; no liability shift
DCAPVisa Digital Commerce Authentication Program — structured data sharing for interchange benefits
TRATransaction Risk Analysis — SCA exemption based on acquirer's rolling fraud rate
ECIElectronic Commerce Indicator — code returned by card network indicating authentication method used
Challenge IndicatorthreeDSRequestorChallengeInd — field signalling whether merchant requests or waives a challenge
Authorization BoostStripe's retry mechanism for declined transactions using 3DS

Benchmarks (as-of 2026-10-23)

MetricValueSourceNotes
Global frictionless avg~64%Ravelin/Stripe (2024 data)Likely lower in 2026 per Ravelin March 2026
Practical frictionless (complete data)80–85%Checkout Experts (2026-10)
Optimal frictionless (low-risk, high data quality)85–95%PaymentBrief (2026-04)
Countries where rate fell/plateaued (2025→2026)28 of 37 (76%)Ravelin (2026-03-19)"Frictionless paradox"
3DS success rate — UK95%Ravelin (2026-03-19)Success rate ≠ frictionless-only rate
3DS success rate — France91%Ravelin (2026-03-19)Success rate ≠ frictionless-only rate
3DS success rate — US88%Ravelin (2026-03-19)Success rate ≠ frictionless-only rate
Frictionless rate — Japan (post-mandate, April 2025)~60%Stripe Sessions 2026
Japan: dispute rate change post-mandateDown >50%Stripe Sessions 2026
Just Eat Takeaway: conversion uplift (Adyen Authenticate)+12%Quantum Payments (2026-09-24)Eligible traffic only
Too Good To Go: conversion uplift (Adyen Authenticate)+11%Quantum Payments (2026-09-24)
Conversion loss per challenge (Stripe analysis)3–8%PaymentBrief (2026-04-12)vs. frictionless
Conversion loss: well-implemented integration challenge3–4%PaymentBrief (2026-04-12)
Conversion loss: poorly optimised in challenge-heavy markets8–12%PaymentBrief (2026-04-12)
Data Only (Europe, >€250): conversion vs. no data sharing+3.2%Stripe Sessions 2026
Authorization Boost (Stripe 3DS retries) uplift~90 bpsStripe Sessions 2026
Exemption rejection: share of genuine-approved failures27.8%Forter via Checkout Experts (2026-10)Largest single cause
User abandonment: share of genuine-approved failures12.6%Forter via Checkout Experts (2026-10)
Merchants using SCA exemptions (2026)78%Ravelin (2026-01)Up year-on-year
Challenge completion: EU mature banking app markets75–85%Checkout Experts (2026-10)
Challenge completion: high-SMS / older infrastructure markets55–65%Checkout Experts (2026-10)

What practitioners report

A 2022–2023 UK merchant forum thread reports mobile customers redirected to their banking app for a 3DS challenge "often did not return to complete payment, even with items still in the basket." Several merchants echo the same pattern of challenge-drop leading to basket abandonment. (ukbusinessforums.co.uk, 2022–2023)

The above practitioner forum data is from 2022–2023 (early SCA rollout in the UK) and may not reflect current issuer behaviour or banking app UX quality in 2026.

Practitioner consensus (adjacent forums, ~2024): approximately 1–3% of customers failing 3DS is typical; above 5% suggests older bank apps, older browsers, or issuers falling back from 3DS2 to 3DS1.

Ravelin (2026-01) warns that SCA exemption strategies that worked in 2024 are less effective in 2026 as issuers tighten fraud assessments. Visa's VAMP penalties (effective January 2026) are expected to push merchants toward better 3DS data submission.

Research agent · 2026-10-23