The Future of Credit Bureaus is Real-Time
TL;DR: Credit bureaus still update BNPL data monthly, but consumers can rack up debt across multiple providers in seconds. The technology to do this in real-time exists: Tilores built it for Regis24. The main barrier is not capability but BNPL providers choosing to share data, which is now changing.
| Dimension | Current state (incumbents) | What real-time requires |
|---|---|---|
| Update frequency | Monthly batch (e.g. Experian UK BNPL) | Sub-second to seconds (approx. 300 ms with Tilores infrastructure) |
| Core technical challenge | Messy, inconsistent data from multiple providers | Real-time entity resolution to match records to one person |
| Consumer risk | Full BNPL debt cycle invisible within one billing period | Over-indebtedness caught before the next purchase |
| Who has solved it | Incumbents have not invested; market stays on batch | Regis24 (powered by Tilores technology); Infact Systems (UK challenger) |
| Blocker today | Technical complexity + incumbent inertia | BNPL provider appetite to share data |
On this page
- Why are BNPL loans invisible to credit bureaus?
- Why do monthly updates fall short for BNPL consumers?
- How does real-time entity resolution solve data ingestion?
- How did Tilores build real-time credit bureau infrastructure?
- Which challengers are building real-time credit bureau capabilities?
- FAQ: frequently asked questions
Why are BNPL loans invisible to credit bureaus?
There was big news in the credit bureau and Buy Now Pay Later (BNPL) worlds recently, as Apple BNPL announced that they will start sharing credit data globally with Experian.
This is important because BNPL (in the US) is disproportionately used by financially fragile households. BNPL is firstly a very convenient option for most modern consumers but it is also an attractive option for those who would otherwise struggle to access credit.
However, until now, the lack of BNPL data reporting has meant there is a risk that this unreported debt leads to the financially vulnerable becoming over-indebted. If BNPL providers were to share their data with credit bureaus, as traditional finance providers already do, this should help to prevent consumers falling into over-indebtedness by providing BNPL providers with the necessary data to allow them to make responsible, informed lending decisions.
Nevertheless, there is a flaw in this data-sharing plan.
In the UK, where 50% of consumers have used BNPL, Experian has proudly shared that they are now including BNPL data in their UK credit data.
Why do monthly updates fall short for BNPL consumers?
The problem? The frequency with which Experian’s BNPL data is updated: monthly.
The average repayment period of BNPL loans is approximately 90 days. Indeed one of the attractions with BNPL is how easily and quickly, normally at the point of purchase, a consumer can be approved for credit.
Therefore it is pretty easy to rack up excess BNPL debt in a very short period of time if one uses multiple different BNPL providers.
Given the ease and speed with which consumers can rack up BNPL debt, it seems ludicrous that the system that is meant to be there to protect them runs with such a massive time delay.
If the world of finance is now providing credit in real-time, why can’t the credit bureaus update their data in real-time?
How does real-time entity resolution solve the data ingestion challenge?
The truth is that they don’t do it because it is technically very challenging. Especially for the incumbents that are slow to invest in new technology.
The data that comes from retailers and BNPL providers is often inconsistent and “messy”, meaning that matching all relevant BNPL data to a single person is difficult. Especially in real-time.
This is the core entity resolution challenge: linking records from different sources to the same individual, accurately and in real-time, even when the data is inconsistent. Understanding the trade-offs in matching quality (see precision vs. recall) is central to building a system that does not generate false matches or miss genuine ones.
How did Tilores build real-time credit bureau infrastructure for Regis24?
I know this because we built a real-time data ingestion and matching system for the German consumer credit bureau, Regis24, before spinning the technology out into a separate company - Tilores. It took us years to develop and in the end the technology was so advanced that we decided to file a patent. It was not straightforward to build.
The reality is that Regis24 is possibly the only credit bureau in the world that could today act as a real-time credit bureau for BNPL data with little effort as that technology we developed is still the core data infrastructure for the company.
I think this is going to change pretty soon. Monthly updating of credit bureau data is not fit for the modern finance world. If I can access hundreds or thousands of dollars of credit in seconds, there is no good reason why my credit profile should not be updated seconds (or 300ms later if you use Tilores) later.
In Germany, Regis24 could do this, if the BNPL providers had the appetite to share data. Ironically, BNPL is such an old model in Germany (so-called “Kauf auf Rechnung” - literally “purchase by invoice”) that it is much more accepted and hence I feel there is less political pressure for the BNPL providers to submit their data.
Which challengers are building real-time credit bureau capabilities?
In the UK and US that is not the case. The pressure is on the BNPL providers to share their data to protect consumers.
We are working with one stealth company to build a real-time credit reporting system for BNPL providers in the US, so the appetite is clearly there. They have a number of US BNPL companies interested in their proposition.
Meanwhile in the UK, perhaps challenger credit bureau providers like ClearScore or Infact Systems (actually founded by ex-Experian techies) could be the future of credit reporting. Infact certainly highlights their real-time capabilities, but they need BNPL companies to be willing to share data with them.
The historical blocker to starting a challenger credit bureau has always been access to data - but if these challengers have technology similar to Tilores that allows high-scale, real-time data ingestion and updating of credit profiles, then BNPL could be just the niche the new credit bureaus need to open up a new market and finally disrupt the incumbents.
See what resolved entity data does for your business — and your AI.
See it on your own data: book a demo for a walkthrough on your records, or get the evaluation build to try resolved entity data locally.
FAQ
Why do credit bureaus still update data monthly instead of in real-time?
Monthly batch updates are a legacy of how data has historically been collected and processed from lenders. Real-time ingestion is technically very challenging: data from retailers and BNPL providers is often inconsistent and messy, meaning that matching all relevant data to a single person in real-time requires sophisticated entity resolution infrastructure. Incumbent bureaus have been slow to invest in the new technology needed to overcome this.
What is the risk of monthly credit bureau updates for BNPL consumers?
The average BNPL repayment period is approximately 90 days, and consumers can be approved for BNPL credit in seconds at the point of purchase. Monthly updates mean a consumer can use multiple BNPL providers to accumulate significant debt within a single billing cycle before any bureau sees it, creating over-indebtedness risk, particularly for financially fragile households.
How did Tilores solve real-time data ingestion for a credit bureau?
Tilores built a real-time data ingestion and matching system for Regis24, the German consumer credit bureau, before spinning the technology out into a separate company. The system resolves messy, inconsistent incoming records to a single person in real-time (Tilores can update a credit profile approximately 300 ms after a new data event) and the technology was advanced enough for the team to file a patent on it.
Which credit bureaus currently have real-time data capabilities?
Regis24 in Germany is described as possibly the only credit bureau in the world that could act as a real-time credit bureau for BNPL data today, because its core data infrastructure was built on the real-time ingestion and matching technology that later became Tilores. Challenger bureaus like Infact Systems in the UK also highlight real-time capabilities, but depend on BNPL providers choosing to share data with them.
What does real-time entity resolution have to do with credit bureau data quality?
The core technical problem in real-time credit bureau data is entity resolution: matching inconsistent, messy records from multiple BNPL and retail providers to the same individual person, in real-time, without errors. This is the same entity resolution challenge Tilores addresses across financial services. Tilores resolves records at ingestion so that a complete, accurate customer profile is always available at query time.
See what resolved entity data does for your business — and your AI.