Veriff found attempted fraud in 19.2% of ecommerce verifications included in its 2026 report , compiled from 2025 customer data. The figure describes pressure within those observed sessions. For a company considering verification, it opens a second question: how many legitimate customers can pass without admitting the risk it wants to avoid.
Selection has two possible costs. Approving an impostor exposes the business to losses; rejecting or losing a valid customer removes a commercial relationship. Service value lies in managing those outcomes together. A detected-attempt rate cannot calculate it on its own.
An attempt rate is not a loss avoided
The same report's global aggregate was 4.18%. The difference places ecommerce at a high exposure within the supplier's data. The aggregate includes sectors, so the bars compare different observed scopes. Neither represents population-wide fraud prevalence or a share of money lost.
Veriff 2025 customer data: attempted fraud by observed scope
Net fraud as a percentage of verification attempts in the 2026 report. The aggregate and ecommerce subset are different scopes, not independent samples.
% of verification attempts
Tap or hover for details. All values are also available in the data table.
Collapse notes
Vendor customer sessions, not population prevalence or financial loss. The aggregate includes sectors. Raw subgroup sizes and uncertainty are not supplied here.
Explore the data table
| Category | Net attempted-fraud rate |
|---|---|
| Global aggregate | 4.18 % of verification attempts |
| Ecommerce subset | 19.20 % of verification attempts |
Attempts also carry different economic consequences. Opening a false account, accessing an existing one and completing a high-value operation involve different sums and tasks. Turning risk into a buying decision requires identifying the event verification protects and what follows its approval.
Identity is one risk within a broader set. The MRC and partners’ survey, reproduced by Statista and conducted in late 2025, found that 31% of fraud managers had experienced identity theft and 24%, account takeover, while 41% cited refund or policy abuse. These are shares of merchants reporting an experience, not fraudulent-transaction rates. The comparison separates risks that may require different controls and avoids promising that an initial identity check resolves every form of subsequent fraud.
Merchants report risks at several stages
Selected fraud types experienced in the preceding twelve months. Global survey fielded November–December 2025.
Tap or hover for details. All values are also available in the data table.
Collapse notes
Statista / MRC: 634 ecommerce fraud managers. Business experiences, not shares of transactions or losses. Selected non-exclusive types. Market context, not Veriff effectiveness. The larger primary payments-and-fraud study total does not replace this subgroup base.
Explore the data table
| Category | Respondents |
|---|---|
| Refund or policy abuse | 41% |
| Real-time payment fraud | 38% |
| Identity theft | 31% |
| Account takeover | 24% |
That difference matters to Veriff's commercial promise. Sector data can justify a discussion about exposure, but customers need to assess the solution within their own workflow and rules. Transferring the report percentage into projected savings would exclude the value of each operation and the buyer's other defences.
The legitimate customer belongs in the result
Veriff's public platform distinguishes initial identity checks, biometric authentication for returning users and business verification. That division places controls at different decisions. Asking for proof at account opening and confirming who returns serve different commercial purposes.
Evaluation needs to follow the complete journey. A user can begin capture, repeat it and leave before receiving a decision. Another can be declined and recover access following review. Reducing that journey to automated approvals loses information about the experience and the assistance required.
The cost of excluding a valid person also appears beyond document checks. Forter and Opinion Matters’ 2023 survey, reproduced by Statista, records false payment declines among 56% of US and 35% of UK shoppers surveyed. Forter retrospectively confirms both figures in its 2024 report. This is a reported payment experience, not an identity-verification error rate or a Veriff result. Its role in the comparison is to show why correctly admitting customers needs its own record alongside blocked fraud.
False rejection also affects legitimate buyers
People reporting an incorrectly declined online payment. Two markets from the 2023 survey.
Tap or hover for details. All values are also available in the data table.
Collapse notes
Forter / Opinion Matters: 5,000 adults across five countries, at least two active ecommerce accounts and an online purchase in the previous three months; per-country base not verified. Historical context. Share of people, not transactions; not KYC, document-check abandonment or Veriff errors.
Explore the data table
| Category | Respondents |
|---|---|
| United States | 56% |
| United Kingdom | 35% |
| Observation | Economic question |
|---|---|
| Fraudulent attempt declined | Which operation is protected and which potential loss remains exposed |
| Legitimate customer completes the process | Which commercial relationship can begin or continue |
| Retry, abandonment or review | Which work it adds and which customer can remain unserved |
Collapse notes
A useful comparison keeps the customer type and protected event stable. If a group supplies harder-to-check documents, its decision time has a different meaning. Speed is valuable when it accompanies an appropriate decision and an operation the customer can complete.
Cost also continues outside the verification engine. Retries, review and assistance for blocked users require work from the buyer. A lower session price can accompany more effort per admitted account. The economic comparison needs to follow that outcome alongside residual risk.
Expansion needs another reason for use
A company buying initial verification may later need to protect access or check business partners. Platform breadth opens those conversations, but each module adds a task and a cost. Another available product does not establish that the account needs to buy it.
Expansion earns its place when a specific event remains outside the previous control and the buyer needs to address it. Its consequence supplies the criterion: which access it protects, which relationship it enables and how much work it adds. That explanation lets the sale follow an observable need rather than a general trust label.
Veriff's strongest commercial argument connects what blocked fraud protects with what a legitimate customer can do. An admitted account completing its intended operation provides economic evidence closer to the business than a fast verification whose journey ends in abandonment.
In this article
Marian Vasile
Marian Vasile is a MarTech and growth professional. His work spans product marketing, SEO, paid advertising and deep online tracking, including behavioral analytics and attribution.
