At the July meeting of the Mid-Market Tech Forum, attendees saw analytics in action before breaking into groups to discuss the scope and use of analytics in different parts of their business. The FRC also provided an initial view of findings from its upcoming thematic review, which suggests a sharp uptick in the use of data analytics in audit, certainly at the top end.
Analytics in action
The NAO introduced its impressive use of machine learning and data analytics within its audits (the NAO undertakes over 400 financial audits, including some of the largest ledger implementations in the UK). The long term vision is for analytics that are integrated, scalable and reproducible. Examples shown included a very simple and visual GL analysis, text analysis using natural language processing, and ‘machine vision’, automating sample testing using OCR. There was also a discussion comparing the relative merits of unsupervised versus supervised machine learning.
Ecovis Wingrave Yeats then ran through its use of data analytics, showing real time dashboard analytics outlining the firm’s productivity (and which timesheets needs to be completed), as well as working capital by partner. A particular point of interest was a graphic outlining a network of potential client opportunities, linking internal and Companies House data.
Actual and potential use
The Forum then broke into separate sessions to consider the actual and potential use of data analytics in audit, advisory and internal, and what skills were needed as a result. The most popular breakout discussed audit; highlights included the variety of uses currently, more work expected in relation to fraud, the importance of using the correct terminology (100% data analysis, not 100% testing), and the issues of how you provide evidence and report on machine learning.
The group looking at advisory saw opportunities with forensics, corporate finance and tax, as well as spotting and advising on other areas where clients are struggling (such as a profit decline). The group also discussed the increasing availability of machine learning objects which can help jump start development in this area.
The group considering internal use noted that data analytics should be used on current data before you start, as data is often ‘all over the place’. It can be used to improve various current processes (such as marketing, web traffic and social media). With client work, data can be combined from multiple sources to help benchmark performance.
And the group looking at skills considered what was needed at different levels, with a realisation that at junior levels more technical and data skills were needed, including statistics and programming. At the managerial levels, skills needed include analytics understanding, interpretation and communication.
Future sessions of the Forum will discuss the skills requirements in greater detail (who you need in your team) as well as drilling down further into use cases for AI and analytics (including real examples presented by the participants). If you are interested in joining the group or discussing the findings in more detail, please comment below or contact me direct. I will be writing up a fuller version of the Forum for publishing in the Tech Faculty magazine Chartech. The Forum next meets in early October.