20 Apr Uplift: Converting More Leads to Loans
By Noah Fitzgerald, CPP
Lenders are missing significant opportunities: There are good borrowers that are declined and there are approved applicants that never originate. Considerable time and money is spent on finding and buying leads that are not converted. Declining good business results in lost revenue and loss of a good client. Approximately 5 – 15% of all approvals never convert to an origination and 30-60% of all applicants are declined. These problems exist due to the lack of data available on the borrower.
Traditional data on the borrower places them into a decline or a higher risk profile. Lenders’ scoring and risk models put the borrower outside of approval guidelines. So, how can a lender convert more of the approved apps and safely approve more of the declines?
Finance companies need data to make decisions on approving a customer for credit and determining rate and term. They use data from multiple sources to ascertain an individual’s ability, intent and willingness to repay before making those decisions. The applicants that are approved are made an offer based on their risk profile with specific terms and rates. Those that are declined, are told why and that’s that. But that is not the end of the story, as in most companies 5% of approved applicants never originate and 70% of the declined applicants are getting approved elsewhere.
The key to maximizing a finance business is all about making the right decision and right offering for each applicant. The key to making the right decision is based on the quantity and quality of data available. Traditionally, lenders are getting data from credit bureaus and other alternative sources. Finance companies compile the data for each applicant and then build a model based on the appetite of risk they have. The problem is that the data they are utilizing does not tell the full story for each applicant.
Traditional data sources, like the major bureaus and alternative bureaus, provide scores and data attributes that are from 30 days to 10 years out. This type of data is very predictive for long term loan products such as an auto or home loan. Additionally, this data works only if the consumer has information reported to those specific bureaus. So, if a lender is basing decisions on an applicant that does not have data available from those sources and thus declines them, does that indicate that they would not be a good borrower? No, of course not. This is a typical scenario for millennials, people new to country and many well-off individuals that simply don’t use credit (baby boomers, etc.).
Another scenario to consider: A consumer has been struggling financially for years, filed bankruptcy 2 years ago, the rest of their credit profile does not look good, and they have no new credit since the bankruptcy. This case, would most likely receive a hard decline on first review. What the data does not reveal is that this consumer got a new job three months ago, making $5,000 a month, and has a total expense of $2,200 per month and their income has been trending upwards. Though this sill may not be the best risk profile, it would be good option for a short-term loan product.
The data leveraged from static sources, such as bureaus, is not providing a complete picture on a consumer. There is missing data or a gap in available data to make decisions on seemingly questionable applicants. This leads back to the original question, how can a lender safely approve more applicants?
Some lenders resort to asking those applicants to provide current bank statements. Typically, a difficult process for a consumer to undertake as it requires contacting the bank and having them emailed or logging into the bank online portal and downloading them to then email to the lender. Then the difficulty for the lender to review the statements, even read them if they were scanned or faxed, verify they are not fraudulent, manually calculate and verify income and expenses, etc. This is not only a cumbersome, time consuming process but very expensive and leaves much room for manual error.
However, there is another option, fintech or financial technology, such as Merchant Boost’s Bank Boost, a solution that allows a consumer to connect their bank account to a loan application through online banking connectivity. Once a consumer establishes that connection, Bank Boost then instantly scans up to two years of the consumers banking activity. Every single transaction is read, identified and categorized within seconds. The fintech then provides structured and useable data to the lender, for example:
- Available Balance
- Current Balance
- Income Frequency
- Income Source(s)
- Income trend
- Expenses by category
- Expense Trends
Since the fintech is looking at the full history of the account activity, it can identify the best date for collection and what the estimated balance will be on that specific date.
Looking back at the previous scenarios of millennials, new to country and other client segments, you can see how this technology and access to live data can help validate a consumer’s ability to repay a loan, potentially turning a decline into an approval. Merchant Boost has performed various validation studies with multiple lenders offering various loan services and the results are what you would expect. Bank Boost data allows lenders to more finely identify customer segments that significantly outperform traditional risk models leveraging bureaus and other alternative sources. If fact, one study revealed that Bank Boost was better able to predict First Payment Default of high risk installment loans than that of a combination of 13 other alternative data sources and tools. Furthermore, the lender reduced the number of bad loans and increased the number of good loans by a significant difference, leading to an effective lift of 5% in good originations.
There is simply no denying how impactful this data is. Where other data sources are pulling from databases or static sources of aged data, Bank Boost is pulling “Live” data that is seconds old. The relevance of this, from an underwriting and risk perspective, is critical, not only to validate traditionally sourced data, but to better identify the full picture of a consumer’s financial position as of right NOW. This insight provides underwriters with actionable intelligence to more accurately verify a consumer’s ABILITY to repay a loan. Additionally, because the data is all pulled and compiled electronically, it can be folded into automated workflows without requiring costly manpower to review it.