We Counted Which Apps Send OTP Codes to Shared Virtual Numbers in 900 Trials

We Counted Which Apps Send OTP Codes to Shared Virtual Numbers in 900 Trials

The reliability of online verification systems has long remained a subject of frustration for developers, quality assurance professionals, and privacy-conscious internet users attempting to register for digital services using temporary or shared virtual phone numbers. To shed light on the mechanics of SMS-based two-factor authentication (2FA) and one-time password (OTP) delivery, the operations team behind the online SMS reception service Receive SMS conducted an extensive empirical study. Spanning a three-week period from October 27 to November 16, 2026, the investigation analyzed approximately 900 distinct trials involving free shared numbers from a single operational pool. The study tracked whether these numbers successfully received automated authentication codes or suffered complete delivery timeouts, categorizing the findings meticulously by application and geographic region.

Background and Context of the Testing Period

The necessity for virtual phone number testing stems from the modern digital ecosystem’s heavy reliance on mobile-based identity verification. Across the globe, platforms ranging from social media networks and financial technology applications to e-commerce marketplaces and developer utilities mandate phone number verification to mitigate bot registration, prevent financial fraud, and comply with regulatory KYC (Know Your Customer) standards.

However, software developers conducting cross-border quality assurance (QA) testing, penetration testers evaluating registration flows, and everyday users prioritizing personal data privacy frequently rely on public, shared virtual numbers. These users often encounter scenarios where a verification code fails to materialize, leaving them uncertain whether the failure originates from the platform’s SMS gateway provider, the local telecommunications infrastructure, or deliberate filtering mechanisms implemented by the application itself.

To address these ambiguities, the researchers isolated a controlled sample of 900 trials. A "try" was defined strictly as a verification attempt utilizing a free shared number that either successfully captured the incoming security code or exhausted the maximum allowed waiting period without receiving any message. To maintain the statistical integrity of the dataset, instances where users prematurely abandoned or manually recycled numbers were excluded from the final tabulation, ensuring that the results reflected system-level delivery performance rather than user impatience. Furthermore, to prevent statistical anomalies skewing the results, the published metrics focused exclusively on application and country pairs that recorded a minimum threshold of eight individual trials.

Granular Breakdown: Apps That Usually Deliver Versus Those That Fail

The empirical data revealed a stark division in how different platforms interact with shared virtual number pools. Certain applications demonstrated remarkably high delivery success rates, indicating either a more permissive SMS gateway configuration or a lower threshold for flagging virtual carrier networks.

At the forefront of reliable delivery, the French marketplace platform Marktplaats achieved a staggering 98 percent success rate, successfully capturing 45 codes out of 46 recorded trials when utilizing French-allocated virtual numbers. Major global utilities also showed varying degrees of compatibility depending on the geographic jurisdiction. For instance, YouTube registrations utilizing Pakistani virtual numbers achieved an 81 percent success rate (13 codes received across 16 trials). Meanwhile, artificial intelligence research firm OpenAI demonstrated a 70 percent success rate (19 codes out of 27 trials) when tested via Indian virtual numbers.

Other notable performers included Indian telecommunications provider Airtel and Nigerian fintech platform Opay, both achieving a 67 percent success rate (14 codes out of 21 trials respectively). Freelancing marketplace Fiverr and community chat platform Discord, both evaluated using German virtual numbers, tied at a 62 percent success rate, securing 16 of 26 trials and 8 of 13 trials respectively.

Conversely, the study identified a significant cohort of applications that exhibit near-zero delivery success when interacting with shared virtual infrastructure. Messaging giant Telegram emerged as one of the least compatible platforms for shared numbers across multiple regions. When tested via Indian virtual numbers, Telegram successfully delivered only 2 codes out of 39 attempts, resulting in a dismal 5 percent success rate. The performance deteriorated further in Nigeria, where only 1 code was delivered across 13 trials (8 percent), and dropped to a complete 0 percent success rate in Pakistan, where zero codes were received across 11 trials.

Other underperforming pairs included YouTube registrations routed through Philippine virtual numbers, which yielded a mere 13 percent success rate (4 codes out of 31 tries), and food delivery service Foodpanda in Pakistan, which recorded a 0 percent success rate across 8 trials.

We counted which apps send OTP codes to shared virtual numbers (900 tries, Nov 2026)

Surprising Findings: Geography, Algorithmic Filtering, and User Behavior

The analysis uncovered three critical phenomena that challenge common assumptions held by developers and consumers utilizing virtual SMS services.

First, the research established that geographic jurisdiction plays as critical a role in SMS delivery success as the target application itself. An application that functions smoothly in one country may fail entirely in another due to regional variations in telecommunications routing, local aggregator agreements, and anti-fraud filters. When aggregated by country, France emerged as the most reliable region in the dataset, with virtual numbers successfully receiving codes on nearly 90 percent of all attempts. Germany and Nigeria followed, each achieving success rates exceeding 40 percent overall. India hovered around a 33 percent success rate, while Mexico and the Philippines demonstrated the lowest regional reliability, frequently resulting in complete delivery timeouts.

Second, the data confirmed that major digital platforms actively and systematically restrict or entirely refuse delivery to known shared virtual number ranges. The structural refusal of messages to platforms like Telegram—which delivered a meager 8 total codes out of 127 cumulative trials across all tested countries—and Amazon, which recorded only 2 successful deliveries out of 50 attempts in broader exploratory logs, indicates sophisticated carrier-intelligence checks. Modern applications frequently cross-reference incoming number requests against databases of known Voice over IP (VoIP), virtual mobile network operators (VMNOs), and shared SMS endpoints. When a match is detected, the application’s backend often simulates a successful send or silently drops the message without ever dispatching an SMS payload to the gateway. Recognizing this, the research team permanently removed these incompatible applications from their public free number lists to prevent users from engaging in futile verification loops.

Third, the study highlighted a widespread behavioral flaw among end-users: premature abandonment. A substantial proportion of tested numbers were manually cancelled or discarded by users approximately 90 seconds after issuance. International telecommunications routing, particularly when traversing multiple carrier handoffs and anti-spam inspection layers, frequently introduces latency that stretches delivery times to several minutes. The researchers emphasized that waiting the full allocated window—typically between 10 to 15 minutes—is often necessary before declaring a verification attempt failed.

Implications for Software Quality Assurance and Security Testing

For software engineers, quality assurance managers, and cybersecurity professionals, these findings carry profound operational implications. Automated testing frameworks that rely on free, public virtual numbers for continuous integration (CI) pipelines or end-to-end (E2E) testing of sign-up flows face severe limitations. Because platforms continuously update their anti-fraud heuristics and blacklist virtual carrier blocks, public SMS reception endpoints are inherently unstable testing assets.

Enterprises conducting legitimate compliance testing, regional accessibility checks, or automated QA routines are increasingly advised to transition away from public, shared virtual number pools. Instead, industry standards point toward dedicated, private virtual numbers leased from reputable telecommunications providers, or utilizing developer-tier Application Programming Interfaces (APIs) provided directly by verified mobile carriers. These dedicated assets provide predictable routing, maintain lower reputation risks, and drastically reduce the incidence of false-negative delivery failures during automated test cycles.

Open Data and Availability of Research Metrics

In the interest of fostering transparency and enabling further academic and industrial research into telecommunications security and OTP delivery mechanics, the complete dataset compiled during the study has been made publicly available.

The comprehensive dataset, encompassing 132 distinct application and country pairs represented strictly through aggregated statistical counts devoid of any personally identifiable user data, phone numbers, or message content, has been published on Kaggle under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Researchers, developers, and data analysts can access the raw metrics via the repository titled "SMS code delivery to free virtual numbers by app." Additionally, the research team developed an interactive data visualization dashboard hosted on Observable, allowing technical professionals to explore cross-tabulations of success rates, geographic distributions, and application-specific delivery trends in real time.

As digital identity verification mechanisms continue to evolve in response to automated threats, empirical studies of this nature provide vital visibility into the hidden friction points that govern global telecommunications infrastructure.

Comments

No comments yet. Why don’t you start the discussion?