The impact of digital business assessment tools on entrepreneurship: experimental evidence from FinTech clients in Brazil

How to cite

Cossio, Diego Vera; Arraiz, Irani; Scartascini, Carlos; Lanzalot, Maria Laura; Valle, Emiliano Diaz Del; Parrilli, Cristina. 2024. "The impact of digital business assessment tools on entrepreneurship: experimental evidence from FinTech clients in Brazil." AEA RCT Registry. AEARCTR-0014011. https://doi.org/10.1257/rct.14011-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Diego Vera Cossio* , Irani Arraiz , Carlos Scartascini , Maria Laura Lanzalot , Emiliano Diaz Del Valle , Cristina Parrilli

* Primary investigator

Registry ID
AEARCTR-0014011
DOI
https://doi.org/10.1257/rct.14011-1.0
First registered
Last updated
Status
In development
Trial dates
2024-08-20 – 2025-12-31
Countries
Brazil ()
Registry record
http://www.socialscienceregistry.org/trials/14011

Abstract

We study the impacts of a business diagnostictool delivered to entrepreneur clients or potential clients of one of Brazil's largest FinTech lender. The intervention consists of an online business assessment tool that identifies areas for improvement, and provides entrepreneurs with educational videos that target those areas of improvement. We use a waitlist design to implement this intervention. The treatment group completes the tool earlier while the control group will complete six months later. We aim at studying the effects of the intervention on two set of outcomes. First, we will focus on the impacts on business practices. Second, we will study the impacts on loan repayment, take up and approval of new loans, and credit scores. The intervention will be implemented on sample of entrepreneurs who are either current clients of the partner financial institution (FI), former clients of the financial institution, or individuals that applied for a loan but were denied a loan from the partner FI. We will analyze heterogeneous effects across these three groups, and across other key dimensions such as gender, age, economic sector, ex-ante credit worthiness, and length of their relationship with the lender.

Intervention

The intervention consists of an online business diagnostic tool (BDT) that identifies areas for improvement, and provides entrepreneurs with educational videos that target those areas of improvement.

Individuals assigned to the treatment group will receive an email invitatiion from the partner FI to complete the BDT online. Upon completion of the BDT, participants will obtain an score and will be offered taylored recommendations based on the areas that the BDT identifies as needing improvement. The recommendations will be delivered through short videos.

Experimental design

Waitlist design with a pure control group.

1.Intervention group. This group will receive a personalized invitation and ongoing encouragement to utilize the tool at the beginning of the evaluation period.
2. Waitlist control (delayed intervention group). This group will receive the invitation to engage with the intervention after a delay period of 6 months to allow for the impact of the business diagnostic tool to be felt.
3. Pure control. This group will not receive an invitation to use the business diagnostic tool. Instead, we will collect only administrative data on this group, to compare their business success to those who received the tool. However, we expect that some members of this group will utilize the business diagnostic tool if and when it is made available to all Creditas customers, and we will monitor this group's uptake of the treatment to determine (a) the ‘natural’ rate of takeup of the product; and (b) determinants of takeup.

We choose to divide the groups in this way because of a fundamental measurement issue: the self-reported outcome measures are generated through the process of being treated. The pure control group enables us to measure outcomes in the longer term that will be measured with administrative records (debt repayment, loan uptake, and information in the credit bureau).

Design details

Randomization method
randomization done in office by computer. Stratified by two groups: 1) current clients. 2) previous clients or individuals who applied for a loan but whose application was previously denied. Within each group we will stratify by gender and region (state).
Randomization unit
individual-level randomization
Clusters
Sample 1: current clients 14,000 individuals Sample 2: prospective clients 283,000 individuals
Observations
297,000 individuals
Arms
Sample is equally split across a treatment group, a waitlist control group, and a pure control group
Primary outcomes
self-reported business practices related to the 3 pillars targeted by the BDT: business basics and separation of personal from business accounts, financial planning, and business financing.
Secondary outcomes
Administrative data: debt repayment, applications for new loans, take-up of new loans, terms of new loans, and credit scores. For all those in the target populations who recently applied for a loan and met pre-qualification requirements (including all current customers), we are able to access a comprehensive credit report. This includes both a numeric risk score as well as data on whether the customer has missed loan payments or is over-indebted. ○ For current customers, we are able to observe loan repayment data.

Cited by

No citation data available for this registration yet.

Keywords