Data integrity: the holy grail of operational efficiency
By Douglas Greenwell, Global Head of Sales.
While those in the field of medicine are chasing a vaccine for Covid-19, firms elsewhere are seeking a different holy grail: operational efficiency in extraordinary times.
The pandemic has created an unprecedented environment for businesses, forced to adapt overnight to remote working and accelerate their pace of digitalisation. A recent report from the International Data Corporation predicts firms will invest $6.8 trillion in digital transformation between now and 2023. Flexible, agile businesses have been able to operate far more easily in current conditions than those still reliant on manual tasks and outdated technology.
A major pain point has been exposed in finance and operations, where spreadsheets and legacy technology are far more prevalent than in customer-facing systems and front-office processes. According to Aite Group, 47% of firms are still processing data manually in these functions.
As we enter a winter of uncertainty faced with economic turbulence, the need to tighten resilience, boost efficiency and reduce costs through improved data accuracy, slicker processes and improved automation has never been greater.
A decade of data
We are living in the ‘decade of data’, with businesses dealing with trillions of lines of data created from a multitude of sources every day. While most firms will have legacy systems in place to deal with a large percentage of this data, anything complex or bespoke will usually need some form of manual processing.
The recent track and trace database disaster was a timely reminder of the risk of depending on Excel, with its wide margins for error and unsuitability for reliably processing data at volume. Accounting teams typically spend time waiting for complex calculations to run across tens of worksheets, often using inherited processes built over years by multiple owners.
The expected trajectory of data explosion means that continuing with such high risk and labour-intensive processes is simply not sustainable. Switching to a more intelligent way of working is a question of ‘when’, not ‘if’. The exceptional circumstances presented by the pandemic has created a stepchange in mindset – automation needs to happen at a far greater rate than ever before.
The pressure is on finance and operations teams to modernise middle- and back-office processes and boost both resilience and business continuity through better data management. Firms are actively seeking ways to make complex data easier to manage and transform systems to ensure data accuracy and a single data truth.
Reconciliation is an important part of this process and an area that we at Duco have seen firms focus on in the past six months. Earlier this year we developed a Reconciliation Maturity Model that identifies a five-step journey organisations can take as they progress from manual reconciliation, through to part-automated, fully-automated and ultimately self-optimising systems that use machine learning to ensure optimum reconciliation efficiency.
The majority of firms sit somewhere in the middle, often in a ‘hybrid’ stage with inflexible technology able to automate some of the simpler tasks but unable to handle anything complex or bespoke. This creates a patchwork of fragmented processes and can result in inconsistencies, errors and duplicate work.
What we’ve seen is that Covid-19 has accelerated the speed at which organisations have wanted to move beyond this – often far faster than the pace that change programmes usually progress at. Firms are now actively seeking ways to make reconciliation easier to manage through modern, flexible systems. The kind that can handle any format of data and be relied upon to solve multiple issues across all roles and functions.
In the past this could have been an arduous process, but businesses now seem much more comfortable with moving reconciliation systems to the cloud. These systems are available on demand (in some cases, just 24 hours), enabling operations and finance teams to set up processes far quicker than on legacy platforms. Cloud-based systems are also far easier to maintain with a remote, distributed workforce.
Maturity through machine learning
Once such modern systems are in place, firms can start to benefit further from having clean digitised data that they can trust. Machine learning technology, for example, is not particularly effective if it is fed poor quality training data, or if a significant amount of that data is still being processed manually.
With a modern, machine learning-enabled reconciliation system in place, organisations can greatly reduce reconciliation setup and investigation times. The system will be able to analyse the data inputs, suggest matching rules, then label and triage any exceptions – all based on past results.
Cloud-based systems have another advantage here. While machine learning technology can be deployed on an on-premise system, the amount of training data is limited to just that one organisation. However, a cloud-based system can learn from industry-wide data and make suggestions based on best practices. This is similar to a traffic app giving you suggestions based on all the journeys its users have made, rather than just the journeys you have made yourself.
As outlined in the Reconciliation Maturity Model, the end goal for reconciliation is a self-optimising model, where every part of the process is automated and the system is able spot data quality issues at source. This removes the need for internal reconciliation, leads to significant cost reductions and frees up teams to spend their time on higher value tasks and projects rather than inefficient processes.
While the technology required for this utopian future isn’t quite ready yet, firms are starting to recognise the steps needed to get there. The transition to machine learning may not happen overnight, but the shift away from spreadsheets to flexible, cloud-based software is taking place faster than ever before.
For more information download the Reconciliation Maturity Model