In Internet financial products, how to determine the user's quota, what indicators should be referenced, and what information should be understood? This article raises several questions about this scenario for your reference.
Someone asked me how the quota should be determined, and they all said that it should be determined according to risk and income, but how to do it exactly? It is easy for laymen to think about this problem simply, and it is easy for insiders to think about this problem as complicated.
You may feel that the high risk and low income will be given a high amount, the high risk and low risk will be given a medium amount, the risk high and low income will be given the lowest amount, and the risk will not be granted
The reason is as simple as this, of course, there is nothing wrong with it.
Then I thought the credit strategy was that simple, but I asked you a few questions.
To be precise, income should actually be the ability to pay debts, and it is also necessary to subtract monthly debts, such as the monthly repayment of car loans and housing loans, as well as the credit card and online loan balances, how to convert them into monthly repayments?
Then how to determine the relationship between the quota and solvency, is it equal? Definitely not, the quota is repaid in installments, is it okay to "solvency x number of installments"?
Interest rates must have been ignored, so let's ignore them for now.
Further, should we consider the quota utilization rate? or even, should we consider leaving a margin factor?
Specifically, the user has 1w of new disposable income per month, do you want him to repay 1w per month, or 5k, or 2w? When we calculate the solvency, we subtract the monthly repayment on the debt side, regardless of the accuracy of the relevant data, there is a more important factor is the influence of the industry.
There are many problems with the accuracy of the calculation of the business impact of peer institutions compared to disposable income.
Suppose a user is a rational person, and he opens three products, a certain bei, a certain article, and a certain flower, and the three product priorities (that is, the user preferences determined by the product and service, including expenditure and fulfillment, etc.) are in this order.
You are now the operator of a certain product, and whether the amount you give is enough is the first question: whether the amount given by a certain bei is enough is the second problem, because when the amount of a certain bei is very high and the user can't use it up, no matter how high you give, it is useless, and the third question is whether a certain flower can raise the amount and upgrade the price to you in front. If you want to be free from others, you have to be the first product, become the first and keep the first. Otherwise, there will be a peer influence, how to consider this peer influence, divide by 1.2 or divide by 2 or divide by 5, or multiply by 2, who knows?
I don't know if there's anything else in particular to keep an eye on.
Based on the above alone, it is very inaccurate to set a quota based on income, not to mention how much of the above information you can get and how much you can't get, and the final result is often that the upper limit of the amount of the amount and the potential performance amount of the user can be accurate by an order of magnitude.
Of course, you can also give a credit enhancement limit based on some credit enhancement information, such as high education, high professional title, etc.
To put it bluntly, the most effective quota strategy is controlled testing. A, B, C, D, and E should be given in the end, and you will know how many groups to measure. You don't even need to cross-divide groups based on risk and income, you can group them based on risk.
If you are not sure, you can use the previous quota strategy based on monthly disposable income as the control group, and divide it into several experimental groups. You can constrain the average amount of the experimental group and the control group to remain unchanged, or you can not need this constraint, as long as your evaluation criteria are complete, in fact, you can measure it however you want.
Of course, the premise is to ensure that the risk is as controllable as possible.
This article is written by Everyone is a Product Manager Author [Lei Shuai], WeChat public account: [Lei Shuai Fast and Slow], original / authorized Published in Everyone is a product manager, without permission, it is forbidden to reprint.
Image from Unsplash, based on the CC0 license.