Your Bank’s Front Door Shouldn’t Keep Good Customers Out
By Jim Marous, Co-Publisher of The Financial Brand, CEO of the Digital Banking Report, and host of the Banking Transformed podcast
Simple Subscribe
Subscribe Now!
We spend a lot of time measuring fraud losses, but far less time measuring the customers we lose while trying to prevent fraud. That imbalance should concern every retail banking executive. Cornerstone Advisors’ 2026 Digital Banking Performance benchmark found that 3.36 digital checking applications were abandoned for every completed application. At the average institution, that represents nearly 9,000 potential accounts that never made it through the process.
Key insight: Identity verification is one of the leading sources of friction, particularly document uploads, yet the answer is not simply to make verification easier. It is to make it smarter. Listen to the full podcast to learn more.
Financial institutions can adjust verification requirements based on the risk presented by each applicant rather than forcing everyone through the same process. The result can be dramatically higher completion without taking on materially more risk.
Need to Know:
- Measure abandonment as a business loss. Fraud is easy to quantify. The value of customers who abandon an application is often invisible.
- Treat identity verification as a risk decision. KYC and CIP requirements establish what information must be collected, but they do not prescribe a single customer journey.
- Use risk signals to determine friction. One credit union increased application completion from 5% to more than 60% by moving document collection out of the standard path and requiring it only when risk warranted it.
- Put acquisition economics next to abandonment data. When institutions spend hundreds or even more than $1,000 to acquire a customer, every unnecessary abandonment deserves scrutiny.
We Have Built a Funnel That Filters
Your welcome mat is out, yet your front door is still locked.
That is how I think about the state of account opening at too many financial institutions. We have designed the front door around controlling risk but haven’t spent nearly as much time asking what that experience is costing us.
The problem becomes even clearer when we look at what happens after someone starts an application. Frida Lebowitz, co-founder and CEO at Debbie, described an industry average application completion rate of roughly 15% on an episode of the Banking Transformed podcast. At some institutions, fewer than five out of every 100 applicants ultimately walk away with an account.
And the biggest drop-off happens at a very specific point: the identity document upload. Frida estimates that roughly 60% of applicants fall away at that step.
That makes the document upload more than a minor annoyance. It is a decision point with a measurable impact on acquisition.
And that acquisition cost is something we can actually put a number on. Nearly half of credit union executives can’t say what they pay to acquire a member. Public financial data puts the average fully loaded cost at roughly $489, with some institutions spending more than $1,000.
Key insight:We know what acquisition costs. We know what fraud costs. We don’t know what unnecessary friction costs.
That is a problem with how we measure the first interaction. We built the process around losses we can count, while the most expensive part of the experience often never appears in a report.
Cornerstone describes digital account opening as functioning more like a filter than a funnel. That’s a useful way to think about what happens when an institution optimizes the process primarily around preventing bad outcomes rather than helping qualified customers complete it.
And this isn’t exclusively a digital problem. Digital account opening represented only about 27% of overall openings in the same benchmark. Most accounts are still opened outside the digital channel, where customers can encounter the same basic friction.
Friction Does Not Equal Security
There is an understandable reason financial institutions have built these processes the way they have. Identity verification is a risk decision, and banks have serious obligations when they open an account. The problem comes when we treat additional friction as the natural consequence of managing that risk.
FICO’s research puts the tension plainly. After surveying 18,000 consumers across 18 countries, the company’s head of fraud and identity made an observation worth taking seriously: friction that drives legitimate customers away often fails to stop determined fraudsters.
Adding another document, another verification step, or another manual review may make a process feel more secure. But if the additional hurdle is applied broadly rather than in response to a specific risk signal, we may simply be making the experience harder for everyone.
The opposite extreme creates its own problem. FICO also points out that streamlining verification without intelligent risk assessment can make it easier for bad actors to get through.
Key insight:The goal isn’t maximum friction or minimum friction. It is appropriate friction.
That means asking a different question at each verification step: What are we seeing that warrants this additional requirement?
Make Risk Determine the Journey
A document upload looks like a friction problem, but I think the more interesting question is why the document is required in the first place.
That’s a risk decision.
And risk decisions can be made differently.
Frida described two credit unions using almost exactly the same technology, yet one had a 5% completion rate while the other exceeded 60%. The higher-performing institution did not simply decide to accept more risk. It changed the customer journey.
It verified identity digitally and moved the document requirement out of the standard path. Applicants were asked for a digital ID only when a risk signal indicated that additional verification was warranted. Same technology. Same regulatory environment. More than 10 times the completion rate.
That should make us ask a different question about every step in account opening:
What risk are we trying to address, and does every customer need the same response?
Frida experienced the problem herself when she applied to one of her partner credit unions and was flagged for potential fraud. Her IP address was in one state, her driver’s license was from another, and her mailing address was in a third.
That sounds suspicious if you are a rule.
It sounds completely ordinary if you are a person.
People move. Driver’s licenses can lag behind reality. Young adults may still receive mail at their parents’ house. Our lives produce data points that don’t always line up neatly, and rules designed around static identities can interpret normal behavior as suspicious.
That doesn’t mean we should ignore those signals. It means we should interpret them intelligently.
The consequences extend beyond a frustrating application experience. About half of Debbie’s users turn to payday lenders, and 89% of those customers return repeatedly. If a payday loan is easy to obtain while a checking account at a bank or credit union is difficult to obtain, we should put those two facts next to each other.
Key insight:The industry spends enormous amounts of energy talking about financial inclusion and attracting consumers into the banking system. We should pay equal attention to whether our own processes are making that system difficult to enter.
When someone says, “KYC won’t let us,” I think we need to ask a better question.
KYC is shorthand. It was never a prescribed customer journey.
For the identity decision we’re discussing, the governing CIP requirements establish the basic identifying information that must be obtained before an account is opened. That is different from saying every applicant must experience the same sequence of screens, document uploads, and verification hurdles.
Bottom line: Better account openings begin with not with less security, but with smarter security that knows when the front door actually needs to be locked.
