Detected flags US$10.4 billion SMB onboarding cost

Detected has published a report estimating that small business merchant onboarding costs the North American market US$10.4 billion a year. Much of that cost, the study argues, stems from information firms must collect directly from applicants.

The report examines onboarding and underwriting for small and medium-sized business merchants in the United States and Canada. It argues that official business registries provide only the data governments choose to collect, leaving financial institutions and payments groups to request much of the remaining information from businesses themselves.

A field-level assessment found that about 57% of the data points needed to onboard and underwrite a North American SMB merchant are available only through direct attestation by the business. These include beneficial ownership, where a business trades, what it sells, expected processing volumes, and whether a person has authority to sign.

Detected describes this divide as the "Collection Ceiling". Information above that ceiling forms an attestation layer that cannot be retrieved independently at scale, even with broader data aggregation or more advanced automation tools.

Policy choices

A central argument in the report is that the gap reflects public policy rather than technical limits. Detected points to differences between jurisdictions, noting that beneficial ownership information is filed and publicly searchable for federally incorporated Canadian companies, while equivalent information is structurally unavailable for Delaware LLCs.

The research also comes as the US beneficial ownership regime has narrowed. Detected said a final rule from the US Treasury removed Corporate Transparency Act reporting requirements for entities formed in the United States and for US persons who are beneficial owners. It said that would leave manual ownership checks embedded in onboarding workflows unless policy or market infrastructure changes.

The estimated annual cost of the attestation layer is US$10.4 billion, with a range of US$3.5 billion to US$31.7 billion. According to the report, the biggest component is not staff time inside financial institutions, but the value tied to abandoned applications when businesses drop out of the process.

That shifts the focus to lost conversion as well as compliance operations. Manual follow-ups, repeated requests for documents or clarifications, and long decision times can all increase the likelihood that a merchant will abandon an application before onboarding is completed.

Limits of AI

The findings challenge the view that better data infrastructure alone can deliver so-called perpetual KYB, or know your business checks that update continuously without repeated outreach to customers. Detected said some pieces of merchant information have no external refresh mechanism and may change without creating any event in a public or commercial database.

Expected transaction volumes, stated business models, and supplier relationships are among the examples cited in the report. In those cases, the information may be correct when supplied, but there may be no independent way to verify later changes at scale.

Detected said artificial intelligence could reduce some of the operational burden involved in collecting this material, but not eliminate the need to ask for it. Autonomous systems may gather information from merchants more quickly, but they cannot retrieve facts that are absent from accessible records.

Navan is cited as an example of onboarding improvement using Detected's system. According to the company, Navan reduced average onboarding time from 52 days to one day and achieved 98% faster onboarding decisions, with support across 99 countries.

The example sits alongside a broader effort by suppliers to banks, payments groups, and fintechs to reduce onboarding friction at a time when compliance requirements remain extensive and customer acquisition costs are under pressure. The report suggests that much of the remaining friction is structural rather than procedural.

The report also includes brief factual context alongside its main findings.

"The industry has spent the last decade trying to make the retrievable half of KYB faster, cheaper and more automated, but nobody has properly measured the information that simply isn't there. You can build a better API, aggregate more sources and deploy increasingly sophisticated AI, but none of those things can retrieve information that was never collected in the first place. That is the Collection Ceiling. Above it, the only source of truth is often the business itself. We need to stop pretending this is a temporary operational inconvenience and start treating it as a structural feature of how SMB onboarding works," said Liam Chennells, Chief Executive Officer of Detected.

This is the first edition of Detected's research into the economics of KYB and merchant onboarding in North America. The report argues that the next stage of onboarding design should separate retrievable data from attestation data, rather than treat every missing field as a sourcing problem.

Detected said organisations track technical measures such as match rates, coverage, and processing times, but often do not quantify the commercial effect of repeated follow-up requests and incomplete applications.

"The most expensive part of merchant onboarding is currently the part nobody measures," said Chennells. "We measure match rates, coverage, API latency and straight-through processing. We don't measure how many times an applicant has to be chased, how many days those requests add or how many businesses give up before they are onboarded. That is the gap this report is designed to expose."