RESEARCH FRAMEWORK | LATAM
LATAM fintech benchmarks built on observed borrower journeys
A trustworthy benchmark makes comparable claims about comparable journeys. FintechProduct focuses on acquisition friction, speed to second loan, CRM intensity, pricing transparency and retention design, using controlled field observations rather than generic rankings.
What a lender benchmark should and should not claim
A public fintech list may rank companies by funding, downloads or market attention. A lending-journey benchmark asks a different question: what effort, cost, information and follow-up did a borrower actually experience under documented conditions?
We do not claim that our selected field studies describe all lenders in Latin America. A valid comparison must state the countries, products, borrower profiles, test dates, loan amounts, terms, study window and any incomplete observations. A scorecard is only as reliable as the evidence collected and the comparability assumptions behind it.
Benchmark dimension 1: acquisition friction
Application friction is evaluated through fields requiring action, time to approval, number of action screens, external documents, observed friction events and biometric or document verification. Each captures a different source of work. Disbursement time is recorded separately from time to approval because a funding-rail delay should not automatically be treated as form friction.
In a controlled 10-lender Mexican study, completion time ranged from 9 to 23 minutes and form requirements from 33 to 60 fields. Those values illustrate dispersion within the tested sample. They are not a verified national percentile or cross-country average.
Benchmark dimension 2: speed to second loan
How many days after first-loan repayment until the borrower sees an active next-loan offer? We log the first observed offer and whether it appears inside the product, via CRM or both. The standard seven-day observation window allows an interpretable bounded result; it cannot prove that the lender has no retention strategy after the window.
This dimension should not be conflated with second-loan application time. A lender can make its repeat application extremely fast while never actively communicating the offer. Activation and execution are two different design choices.
Benchmark dimension 3: CRM intensity
We assess observable communication behavior across the full cycle: channel mix, volume, purpose, timing, personalization, repayment reminders, collections and post-repayment reactivation. Contact frequency alone cannot be read as effectiveness, and an observed regulatory risk should be framed for legal review rather than as a definitive violation.
The communication matrix is particularly valuable because it preserves the evidence behind a summary finding. Teams can separate servicing from selling and trace exactly which messages were received under the study conditions.
Benchmark dimension 4: pricing transparency
The question is not only what the loan costs. It is when and how the borrower sees the total amount, interest, fees, taxes, penalties, extension terms and contract access. A lender can present a public price while revealing economically significant components only near confirmation or later.
Because short-term loans can have unusual fixed-fee structures, comparisons must preserve the term and principal or clearly standardize them. A calculated effective rate, a reported CAT and a contractual nominal rate are not interchangeable numbers.
Benchmark dimension 5: retention design
Retention design evaluates changes from Loan 1 to Loan 2: whether data are reused, identity verification repeats, application time falls, financial terms change and product progression becomes visible. This is a product experience measure, not an estimated borrower lifetime value.
Our observed case studies demonstrate why loan history matters. In one Mexico journey the second application took about two minutes versus 12 minutes 42 seconds initially. In a separate Colombia study it took 3 minutes 41 seconds versus 16 minutes. These examples should not be pooled into a regional average without compatible sampling and study design.
Evidence quality and comparability rules
Each benchmark should specify exactly what was verifiable in the screenshots, contract, message logs and timing record. Inferred events, such as disbursement time estimated from a notification instead of bank confirmation, must remain labeled as inferred. A controlled late-payment phase, if included, also changes how second-loan eligibility should be interpreted.
FintechProduct’s proprietary formulas and calibration remain internal. Public methodology describes measurement categories, not enough operational detail to recreate protected researcher profiles or score weights. Where dimensions cannot be observed, a report should mark the gap instead of imputing values.
How to interpret a scorecard without overclaiming
The output serves executive, Product, Growth, CRM and Risk teams. One number is useful only when paired with the observed events explaining it. A low-friction score can coexist with a late OTP failure, while a good repeat-loan workflow can coexist with no reactivation communication. Strategic interpretation should separate observation, plausible mechanism and decision that the client needs to test.
Explore the Mexico field-evidence overview and Colombia field-evidence overview. To commission a comparable multi-lender study, see the Competitive Benchmark service. This page explains the research logic, not a downloadable public dataset or complete league table.