Time to apply for a loan, Mexican lenders

Before the first data wall

When you review your own lender’s application, start at the first data wall: the first screen that asks for real personal data, an SMS code, bureau consent or an account. In this review, that opening took an estimated 81 seconds at the median and held a median of five fields, a short stretch beside the much longer reach described in lenders’ notices and app permission declarations. For someone using a phone, the time to fill everything before that point was estimated at 22 seconds at Doopla and 4 minutes 22 seconds at Financiera Independencia, whose form came across two screens. These figures come from a phone-use model, not measured times.

Horizontal bar chart of estimated phone completion times before the first data wall for 19 lenders, sorted fastest to slowest. Times range from 22 seconds for Doopla to 4 minutes 22 seconds for Financiera Independencia, across two screens. The median is 81 seconds. Estimates use a phone-use model, not measurements.
For 19 lenders, estimated phone time before the first data wall runs 22 seconds to 4 minutes 22 seconds. Median: 81 seconds.

The routes between those ends varied too: Yotepresto came to about 1 minute 45 seconds, Caja Popular Agustín de Iturbide to about 3 minutes, Alux Capital to about 4 minutes and Aser to about 4 minutes 10 seconds. For your review, the comparison is not just how long the opening takes. It is also what the borrower is asked to provide before reaching the first data wall.

At the 19 lenders where a real form was reached, the first screen held a median of five fields, ranging from two at FinKpital to 18 at Alux Capital. FinKpital’s screen asks for an RFC and a privacy checkbox. Alux Capital’s pre-calificador asks for details including education, dependants, address and employment. Financiera Independencia’s 21-field form came on its second screen, after a simulator, rather than on the first. That distinction matters when you put the opening screen beside the rest of the route.

Side-by-side phone screenshots show FinKpital’s first application screen with 2 fields and Alux Capital’s with 18, the low and high counts among 19 lenders where a form was reached.
Across 19 reachable forms, first screens ranged from 2 fields at FinKpital to 18 at Alux Capital, with a median of 5.

CAT on homepages

The CAT is stated on only some of the homepages reviewed. Of the 94 that loaded, 15 contained the word CAT and eight gave a CAT figure on the homepage itself. That distinction matters when a borrower compares what a lender states before an application begins with what appears later in the application route.

The printed figures span a wide range: Yotepresto showed 26.2%, while Lime24 showed 40,699.7%. They are not like-for-like prices.

Each lender’s figure is its own average or range, for its own product, amount and term, using its own calculation date. The figures were not recalculated on a common loan, so the difference can be reported, but it does not directly compare the price of the same loan.

Logarithmic bar chart of CAT figures printed by 11 named lenders, from 26.2% at Yotepresto to 40,699.7% at Lime24. Credilikeme is shown as its printed range of 83.0% to 3,217.0%.
Printed CAT figures span 26.2% at Yotepresto to 40,699.7% at Lime24, but reflect different products, terms and calculation dates.

Put these checks into a decision flow

Create rules for credit, fraud or collections, then run them on live decisions.

Notices and loaded services

The first data wall does not mark the limit of what lenders’ notices name. In the 93 notices read, keyword counts found biometric data in 36, geolocation in 29 and device identifiers in seven. Fifty-five notices contained a clause on transferring or sharing data with third parties or credit information companies. The same keyword search found Buró de Crédito in 20 notices, Círculo de Crédito and SAT in seven each, RENAPO in six and IMSS in three. These are counts of notices containing the terms, not counts of mentions.

The websites also loaded named services. MetaMap loaded on Yotepresto’s borrower signup page, and Nubarium loaded on Fortrade’s and Sofortaleza’s pages. IP lookup services loaded on Prestadero, Yotepresto, Jeeves and Caja Yanga. Separately, Doopla’s legal pages named MetaMap in a notice. These observations identify services present on the pages or named in a notice. They do not establish why a lender chose a service or what it did with a particular applicant’s information.

What apps declare

The app scan offers a different kind of evidence from a notice: it shows which named services were present in the apps examined, not how a lender used them. The lenders’ sites linked 22 Android apps. Twelve were downloaded and scanned, and 10 downloads failed. Incode appeared in Stori, Klar, Finsus and Fincomún. Belvo appeared in Stori, and AppsFlyer in Stori, Klar and Finsus. Google Play Integrity appeared in Stori, Klar, Finsus, Fincomún and dportenis. Caja Mitras’s app was built on Bankingly. These findings describe what the scan found in the apps it could examine.

A permission declaration tells you what an app may ask to access, not what a borrower is actually asked to grant. Across the 12 scanned apps, precise location and biometric unlock appeared in 10 each, camera in nine and microphone in six. Contacts appeared in four apps, phone state in five, accounts on the phone in three and a list of installed apps in two. Stori declared nine of these permissions, Finsus and Klar eight, and Caja Yanga and Yotepresto one each. The declarations show the possible reach of an app, not whether it requests any particular permission from a borrower.

A grid with one row for each of 12 scanned lender apps and columns for declared permissions. Filled cells mark manifest declarations. Counts: contacts 4, all installed apps 2, phone state 5, phone accounts 3, precise location 10, camera 9, biometric unlock 10, microphone 6.
Precise location and biometric unlock are declared by 10 of 12 apps, but a declaration does not show that an app asks.

Two notices explain why an app might seek contacts. Credilikeme says the permission is requested only for phone top-ups by a user with an active loan. Caja Oblatos says contact access is for buying airtime and paying for a contact. Those stated purposes give a borrower more context than the declaration alone, though the declaration still does not establish that access was requested.

The scan found none of the device and fraud-scoring tools CredoLab, SEON, ThreatMetrix, FingerprintJS, Sardine, JuicyScore, Incognia, Sift, Kount, iovation, BioCatch, ClearSale or Konduto in the 12 apps or on the websites.

The route into the market

The homepage did not always lead a borrower to an online application. Of the 107 lenders, 23 led to an online application, account signup or loan simulator.

Among the 94 homepages that loaded, 63 linked WhatsApp and 20 linked the App Store or Google Play. For a borrower deciding where to begin, the route offered on the homepage is part of the experience, before any application screen appears.

The findings put a short application opening beside a much longer reach in lenders’ notices and app permission declarations. In your own lender’s flow, compare the first screen with the data its notice names and the permissions its app declares.

Put these checks into a decision flow

Create rules for credit, fraud or collections, then run them on live decisions.