Compliance is a conversion problem

David Schreiber

TL;DR

Each follow-up loop in a business onboarding flow loses approximately 15% of customers who have already submitted, and the drop-off compounds with every subsequent request.
In markets where fields are pre-populated from registry data, follow-up requests on those fields drop to near zero; in portfolios without pre-fill, 50% of compliance follow-up questions concern information that could have been retrieved automatically.
Businesses that do not receive an approval or rejection decision within 24 hours activate at materially lower rates than those who do.
Faster, smoother onboarding raises total customer lifetime value (LTV) across the board, not only for high-intent customers, but structurally, across the entire portfolio.

The case for treating onboarding as a revenue problem

Most B2B companies measure their onboarding by whether it satisfies regulators. Few measure it by whether customers actually complete it.

That framing has a cost. Business onboarding — know your customer (KYC) verification, anti-money laundering (AML) checks, identity and verification (ID&V) — is not a back-office formality. It is the first substantive interaction a business customer has with your product. Every unnecessary field, every follow-up request, every broken verification flow is a customer you have already won, quietly leaving.

The companies that recognize this are not just improving their compliance operations. They are improving revenue.

“Compliance is a hygiene factor. When that high bar is met, everything should be about conversion and a delightful customer experience.” — Duco van Lanschot, Co-founder, Duna

The problem is larger than initial onboarding

When companies talk about onboarding conversion, they typically mean the initial sign-up flow. That is the smallest part of the problem.

A business relationship involves many touch points after the first onboarding: product upgrades, periodic reviews, transaction monitoring alerts, reboardings triggered by terms-of-service updates. Each of these is, in practice, another onboarding event. Each has its own completion rate, its own drop-off points, its own operational cost. Most companies have no visibility into any of them.

In our conversations with executives across financial services and regulated industries, the pattern is consistent: organizations can articulate their initial conversion rate, sometimes. They rarely know their reboarding completion rate, their follow-up response rate, or the cost of a single non-response. Without that visibility, there is no baseline from which to improve.

The first step toward better conversion is measurement.

What kills onboarding conversion?

Conversion drop-off is not random. Across the onboarding journeys we see in our portfolio, the same friction points appear with near-universal regularity. Most of them are not caused by complexity in the underlying compliance requirements. They are caused by how those requirements are presented.

Follow-up loops are the single most common cause of mid-funnel leakage. Most onboarding flows ask for the minimum upfront and collect additional information through follow-up requests. Each follow-up loop loses approximately 15% of the customers who have already submitted. That leakage compounds: a second follow-up loses another 15% of the remainder, and so on. The first-time-right metric — collecting everything needed in a single pass — is, for this reason, one of the highest-value improvements any compliance team can make.

A related problem is form design that ignores context. A form that asks a German business to provide its "TIN" — without explanation, without localization, without example — will generate incorrect submissions. A Dutch BV and a German GmbH have different regulatory requirements; a form designed for neither serves both poorly. The result is not just customer frustration. It is a support burden, a delay, and a further follow-up loop.

Beyond form design, state management is a structural failure in most in-house-built journeys. Customers who drop off mid-journey and return to find they must start from the beginning typically do not return. The ability to resume exactly where a customer left off, across devices, removes that failure mode entirely.

Collaboration bottlenecks are the friction point most specific to business — as distinct from consumer — onboarding. The onboarding of a business is rarely the responsibility of one person. Ultimate beneficial owners (UBOs), legal representatives, and multiple signatories may all need to provide information or sign off. Flows that are not designed for this — that route everything through a single login, that have no mechanism to bring in a second party — create waiting, confusion, and drop-off at precisely the moment when the customer is most engaged.

The last major lever is pre-fill, and it is the most underused. Customers are routinely asked to provide information that could be retrieved automatically from company registries or pre-populated from existing CRM data. Every field that can be pre-filled and is not represents additional friction with no corresponding compliance benefit.

In our internal data, 50% of follow-up questions from compliance teams — requests for clarification on submitted information — relate to fields that were not pre-populated. In the countries where we pre-populate those fields, the follow-up rate on those same fields drops to near zero.

What does well-designed onboarding look like?

The gap between a poor onboarding and a good one is rarely a matter of technology investment. It is a matter of design principles applied consistently.

Adaptive journeys collect only what is relevant to this specific customer — their country, their legal form, their business type. They do not ask a sole trader the questions designed for a corporation. Dynamic question logic reduces the total number of inputs required and reduces the surface area for errors.

First-time-right collection means making the risk assessment while the customer is in the interface, not after they submit. If additional information is needed, it is requested in the same session. The follow-up loop is eliminated, not managed.

Multiplayer facilitation treats business onboarding as the multi-party process it actually is. UBOs and legal representatives receive direct, private invitations to provide their portion of the information. The coordinating party — the relationship manager, the compliance officer — has visibility into what has and has not been submitted. Nothing waits in an email thread.

Smart re-engagement surfaces customers who have dropped off and reaches them through the right channel at the right time. In our testing, reminders via SMS and WhatsApp outperform email for re-engagement, particularly for mobile-first customer segments. The lift from well-targeted reminders is typically three to five percentage points.

Where does AI change the arithmetic?

The trends above describe what well-designed onboarding looks like today. AI will accelerate several of them, and introduce capabilities that are not yet standard.

Pre-fill will expand beyond registry data. Rather than populating only what appears in a company registry, AI systems will be able to read a business's public web presence — their website, filings, news coverage — and use that to complete significant portions of the onboarding journey before the customer has typed a single character. For businesses that have previously onboarded with any institution on a shared network, the information is already verified; a new onboarding becomes a consent action, not a data-collection exercise.

Fraud considerations will reshape how identity is verified. As synthetic identity fraud becomes more sophisticated, the reliance on document uploads will give way to biometric checks, mobile push notifications, and multi-party verification flows. Onboarding journeys that are not designed for this now will require significant rearchitecting.

The direction of travel is consistent: fewer steps, more intelligence, faster decisions. The companies building toward this now are not doing it as a compliance exercise. They are doing it because the revenue case is clear.

The Lifetime Value (LTV) Connection

Conversion rate is the metric most commonly cited. It is not the most important one.

What our customers consistently find — and what the data from our portfolio suggests — is that faster, smoother onboarding does not only improve how many customers complete the process. It changes how those customers behave afterwards. Customers have both reported that improvements to their onboarding journeys raised total customer lifetime value (LTV) across their portfolio, not just for high-intent customers, but structurally, across the board.

"It's really a structural shift, the LTV goes up, not just the addressable LTV with conversion rate, but the total LTV actually went up." — David Schreiber, Co-founder, Duna

The mechanism is attention. A business that begins using your product within 24 hours of applying is in a different psychological state than one that waited a week for a decision. The latter has had time to explore alternatives. In markets where the underlying product is commoditized — payments, lending, embedded finance — the experience of getting started is often the differentiator. Customers report that businesses that do not receive an approval or rejection decision within 24 hours activate at materially lower rates.

Compliance built for speed is compliance built for growth. The two are not in tension. Companies that treat onboarding as a revenue function, rather than a regulatory function, are discovering that the economics of both improve.

The practical question

For any organization running onboarding journeys at scale, the starting point is visibility. Before optimizing, you need to know: what are your completion rates at each stage, for each journey type? What is your first-time-right rate? What share of your follow-up requests result in drop-off?