EXCLUSIVE: “Failure is not an option” – Dalbir Sahota, Accuity and Tristan Blampied, Silicon Valley Bank UK in ‘The Paytech Magazine’
With margins in banks’ payments businesses already under pressure, Accuity’s Dalbir Sahota and Tristan Blampied of Silicon Valley Bank UK, explore the impact of failed transactions and how to fix them
The impact of the exponential shift to high-volume, low-value digital payments on banks’ balance sheets over the last few months has been revealed by stark new findings. The True Cost Of Failed Payments Report, based on a survey conducted in early 2021, features responses from more than 200 payments professionals across the banking, financial, fintech and corporate sectors in both established and emerging economies. It pins an eyewatering figure of $118.5billion on failed payments globally in 2020; the total hit incurred in fees, corrective labour and lost business.
Regionally, that figure breaks down to $41.1billion in Europe, the Middle East and Africa (EMEA); $33.7billion in the Americas; and $43.7billion in Asia-Pacific (APAC). While the cost varied globally and by organisation, the average for banks was $360,000 each, and just over $200,000 for corporate businesses. The report defines a failed payment as one that is ‘rejected by a beneficiary or intermediary bank [for] reasons including inaccurate or incomplete information, data entry issues due to human error or poor reference data and validation tools’.
The figures, compiled by Accuity, part of LexisNexis Risk Solutions, reveal how imperative it is for major institutions to grasp the digital nettle and re-orientate their payments businesses to achieve greater efficiency and increase products and services to compete with paytech challengers. But more alarming, perhaps, than the topline financial loss, is the impact of failed payments on customer retention – 60 per cent of institutions surveyed (rising to 80 per cent among those with more than 20,000 failed payments a day) said they had lost customers as a result.
Accuity found that issues with account numbers caused a third of failed payments, with inaccurate beneficiary details making up another third. Manual processes also introduced human error and slowed down payment journeys. Yet, less than 50 per cent of institutions are actively doing something about it. More than a third of ‘payment data elements’ are still validated manually, and two-thirds of organisations find reducing manual processes ‘extremely challenging’.
Dalbir Sahota, global head of KYC and payments product management for Accuity, which specialises in data analytics for risk management and decision-making, says the industry has yet to wake up to the profound impact
of dire payments performance.
“Most organisations do not fully understand the impact, both financially and from a customer retention standpoint,” she says. “Tangible costs like fees and labour might be relatively easy to measure, but the intangible damage, including customer relationships, can be harder to repair.
“The payments market is fiercely competitive, so it is vital for organisations to do more to improve their payments data and reduce failures.” Initiatives like the new ISO 20022 payments messaging standard, the European Payments Initiative (EPI) and the Single Euro Payments Area (SEPA) are all conspiring to help sort out issues with, particularly cross-border, payments efficiency. But, to reap the benefits of this, the report concludes banks have no time to waste in embracing the kinds of digital technology needed to support them.
“The pandemic has increased customers’ expectations of their payments experience. There’s been an explosion of cashless, digital payments and, naturally, as growth happens in payments, there is also a growth in failed payments,” explains Sahota.
Tristan Blampied, director of payments product management at Silicon Valley Bank UK, agrees.“Delay tolerance levels have gone down, so it’s necessary to look at every avenue to validate payments upfront to prevent this,” he says. Yet, as the Accuity survey shows, many financial and corporate institutions are still manually checking and administering payments. “And, where things are done manually, there’s always that chance of something going wrong,” says Sahota.
“Automaton rates for pure domestic payments should be at least in the high 90s. If they’re not, there’s a problem,” continues Blampied. “Cross-border automation is lower and failure rates higher because there are more variables.”
But there is hope, says Blampied: “SEPA is an example of a harmonised standard, using ISO 20022 messaging, being rolled out for participating countries, with positive impact.”
The biggest cost for processors is the manual work involved in operators trying to repair payment instructions, he says.
“Then there’s the notification process – which could also be manual – back to the originator, that the payment has failed. Organisations might need to build expensive, bespoke reports, too, stipulating why payments failed. Then, customers with lots of failed payments are likely to take their business elsewhere, in a competitive payments market with lots of payment provider options.”
Sahota agrees: “Customer churn associated with failed payments is probably the gravest consequence. What the market’s achieved, in domestic payments, is really strong, and it’s a case of mirroring that in cross-border, because, more than ever, we operate in a global world, with supply chains and companies working across borders, relying on payments being made right first time. Many sectors have really suffered because of the pandemic, cashflows are dependent on payments being successful and one not reaching its destination, first time, has a consequence for business performance.”
She continues: “The shift towards businesses using more digital, e-commerce channels, has been phenomenal but, when it comes to implementing that, if organisations haven’t solved the causes of their failed payments, and digitised and automated their checking processes, they’re still going to experience failures and now need to address those underlying problems.”
Happily, there are lots of solutions available to help improve payments automation and increase straight-through processing (STP), adds Blampied: “Rule-based processing is already fairly commonplace in financial institutions, where there’s one field missing from a payment message and predefined rules ascertain what’s missing and add it in, without manual intervention, so that a payment doesn’t fail,” he explains. “We’re now seeing the addition of artificial intelligence (AI) to further automate the process and repair instructions that have multiple layers of issues within them.
“There is only so far rule-based processing can go. Natural language processing (NLP), however, can assign a value to every piece of information within a message and achieve a contextual understanding of what it is, who it’s from, where it’s going and what it’s for. So, using NLP, a system can effectively repair three, four, five different errors within a payment and structure it correctly, before it’s sent through the correspondent banking chain, which is quite amazing.
“Similarly, machine learning is very valuable. It can look at payment repair tasks human operators have undertaken, learn from those and build a model so that, when payments feature those same errors in future, the system can repair them, reducing fraud and risk by removing the need for human intervention.”
“The end outcome, to have payments straight-through process first time, for businesses dealing with high volumes, is super-critical, because it is the bedrock of their business,” adds Sahota. “A one per cent drop in straight-through processing causes blood, sweat, headaches and tears for actors involved in the entire value chain of a payment, from the originator to the beneficiary, to the folks that have to manage all the interactions to get it right. These are all unseen costs, centred around the pain points financial institutions are feeling and, equally, that end-customer pain point, which is very real in the market.
“Reducing failed payment rates requires investment for today and tomorrow, to prepare for what’s coming down the track and not just look at the problems of here and now. It means investing in things like AI that are coming down the line, but also some low-hanging fruits in terms of understanding what datapoints are causing failed payments, and addressing those.
“The demand for faster and competitively-priced payments cross-border, as well as domestic, is very, very critical.” “Automation, automation, automation, and deploying the technologies that can support that, like AI, are vital,” adds Blampied.
The adoption, globally, of ISO 20022 messaging, is also key. “There are proven success cases where that has already happened, such as SEPA, but there is further to go to achieve a harmonised and uniform global standard, to reduce failures and ensure customers get their money on time.”
- The Bank of London Appoints Marc Jenkins as Chief Financial Officer UK Read more
- FF Awards 2022 Highlights Read more
- MAPFRE RE Increases Capital by 250 Million Euros to Offer Greater Reinsurance Capacity to Its Clients and Boost Growth Read more
- Thales Addresses Inclusivity With Its ‘Voice Payment Card’ Read more
- StrideUp Cuts Rates and Launches Its Innovative First-time Buyer Product to the Intermediary Market Read more