Machine Learning Delivers Financial Confidence
Last month Arkk held a roundtable with some of our key tax clients in order to understand the challenges they are facing, how they are progressing with Making Tax Digital (MTD) and what they need to have confidence in the figures they are submitting. Over the next few weeks, we will be sharing these insights.
As we saw in our last blog, our tax clients identified that making stakeholders’ lives easier and reducing risk to the business was key to getting buy in to make changes to archaic tax processes. So why, when it has delivered so much in other parts of the business, has digital transformation stopped before the tax department?
In many cases, tax processes and compliance are still based on software built in the 1990s. So it is uncommon for the tax department to realise the benefits of automation and machine learning.
There are many manual and time-consuming processes in undertaking tax compliance, therefore numerous opportunities to use technology to not only make these processes more efficient but most importantly more accurate.
Automation and Beyond
What was clear from our recent roundtable was that while people may not be using automation or machine learning, everyone is thinking about how it could streamline their business.
As one of the Heads of Tax at the roundtable said, “One of the biggest challenges we are facing at the moment is looking at how to make better use of software. We want to look at how to streamline processes, how to automate things and ultimately how to make [tax compliance] quicker. It seems that everybody just wants reports quicker than they had it last year. So we’re constantly looking at automated solutions.”
When used together, automation and machine learning can massively help to improve business processes. However, this doesn’t mean handing the entire tax compliance process over to technology. Instead, it allows tax departments to save time by not having to do the same repetitive tasks, such as manual adjustments, each month.
Managing The Transaction Volumes
Especially with complex, multi-jurisdiction tax compliance, the transaction volume is so high there is no way for a person to review each entry. Automation and machine learning means that every transaction can be reviewed, anomalies flagged and any interventions taken by the team noted and learnt from for the next report. Over time, the team will need to make fewer interventions and can focus on more strategic tasks.
As one attendee said, “We see AI coming into the process. Effectively the technology is learning what the normality should be. And if any metric is slightly different it will flag it as a red report to the tax team to review. We need this as the volumes are too insane for anyone to actually look at each transaction.”
It’s important to remember that even with machine learning the team still has control. The expertise brought by the tax department shouldn’t be underestimated. Machine learning can’t beat a humans intuition, knowledge and experience to judge why an anomaly has occurred or how it affects a business’s tax position. But it does remove the manual and time-consuming processes that can lead to inadvertent errors.
Fix and Verify – Check For Errors in Seconds
Using machine learning algorithms, such as Arkk’s Fix and Verify capability, to automatically look for duplicate invoices or to interrogate every line-item of data, searching for patterns delivers a level of financial confidence than a human alone can’t. As it learns over time, it becomes faster and more accurate after every reporting period.
One of our clients summed it up nicely, “We are getting a lot of interest from the CFO about actually being solution-driven and technology is a big part of that. The bottom line is if it can be digitised, then digitise it.”
If you are interested in hearing how to digitise the tax department and make the most of machine learning then contact us on 0207 037 2758 or enquiries@arkksolutions.com for a demo.
