Data as a deliverable. Does our thinking regarding audits need to change? Technological innovation in audit and accounting has come in many forms over the last 40 years. The 80s saw the introduction of personal computers. The ’90s saw the potential of the Internet as well as the digitalization of financial records and business process documentation. The mid 2000s and early 2010s saw the widespread proliferation of Software-as-a-Service (SaaS) applications, triggering the creation of more accounting data than has ever been encountered before.
Each of these technological shifts has undoubtedly delivered benefits to audit firm efficiency. Some have created new challenges. An example of this is the proliferation of SaaS management systems that today govern and capture data on client’s business processes. Financial reporting requires consolidation of information from all these disparate systems, whereas audit requires a normalized and deconsolidated view of that same data. Data has become both the problem and the solution.
The audit industry has always embraced new innovations that drive performance, reduce cost, and improve the client experience. AI, machine learning and large language models have arrived in the audit community at a time when auditors and accountants need to manage the vast amount of data created by these SaaS platforms, but to make that work effectively, auditors need to evolve the audit process, and their relationship with both structured and unstructured financial data.
Recreating Paper-Based Processes
The global audit software market was valued at $538 million in 2014. By 2024, that number had more than doubled to $1.27 billion. Looking ahead, the segment is projected to grow to more than $3 billion by 2032.
Auditors use a wide range of tools, applications, and pieces of software throughout the audit
If set up and used correctly, these tools can offer measurable efficiencies.
However, today’s SaaS platforms have simply replicated the outdated paper-based processes of the past and moved them online. Instead of sorting through stacks of paper records in binders and bankers’ boxes, auditors now sort through digital documents on multiple screens. The time savings exist, but there could be so much more.
Further, because many of these client platforms don’t integrate with each other, or normalize the data, auditors are forced to download the data onto a local desktop computer, manually transform and normalize it in Excel, and then reupload a newer version to the shared workspace for the next stage of the audit.
All of this means that auditors are failing to reap the full benefits of digitized data. They are stuck with obsolete, time-consuming, and manual processes. They spend a significant amount of time collecting, reviewing, and normalizing data rather than conducting testing and analysis to validate findings and explore results.
The Juice Wasn’t Worth the Squeeze
The potential for new technologies to transform the audit industry has been known for some time. Consider this quote from a 2013 report by The Association of Chartered Certified Accountants.
“Heading into the 21st century technology trends in cloud, big data, mobile and social collaboration are converging to change the ways in which we consume information technology resources, share knowledge and experiences, and access products and services.”
However, despite large investments in new technologies and tools, auditors have been left unsatisfied with the quality of insights and the impact that data has had on their performance.
For example, previous digital transformation initiatives saw firms attempt to implement data lakes that promised to provide big data analytics and other forms of intelligent analysis.
Many firms loaded terabytes worth of data into data lakes, calling on their most experienced human resources to tag and contextualize their data. However, once the data was manually organized, the data in the data lake would already be out of date, and any new data being added, particularly at the accelerating volumes and speed of how it was being produced by SaaS tools, would and could break the models previously put in place due to the ever-changing business approaches. Without significant periodic manual review and sanity checks by skilled resources, the reports coming from these data lakes quickly lost their value. The effort to build them was not worth the cost.
Today, we don’t have to rely on the manual tagging of experts as we can use automation, machine learning and AI approaches to auto-curate and review data lakes at scale in a timely fashion and at the same pace it is being produced by SaaS platforms. This allows us to again use those experienced human resources in a more efficient fashion to evaluate the system models that govern the data lake curation instead of having them manually curate the lake. It’s the human-in-the-loop approach to data curation today, versus what was the human-is-the-loop approach 10 years ago.
Generative AI as the Catalyst for Data as a Deliverable
The emergence of generative AI (Gen AI) and large language models (LLMs) are the catalysts that are pushing audit firms to adapt traditional audit methodologies and change the way they approach data.
Not since the Internet has a technology had the potential to be so transformative. But while the Internet took decades to become ubiquitous, AI has seen an explosion in use. Globally, the number of people who used AI tools had nearly tripled to 314 million from 2020 to 2024, with projected users growing to 729 million by 2030.
Audit firms are already testing the waters, with roughly two-thirds saying they are considering adding Gen AI technologies to their audit workflow.
For these tools to be effective, they require access to a secure, curated, organized, and auditable data repository that must be up-to-date and readily available. AI must be able to explore the data efficiently by leveraging the platform’s context, indexing, and transformation capabilities. This uniquely calls for the use of auto-curated data lakes governed by intelligent data management platforms.
Transforming Audit Technology for the Future
The past decade has seen the adoption of software tools and applications that have helped to streamline parts of the audit methodology.
However, these applications have continued to rely on the manual, time-consuming, and resource-intensive processes that were common during the paper-based era of the past.
As we look ahead, the firms that embrace a new way of thinking about data management will be best positioned to capture the benefits of artificial intelligence and large language models.
By seeing data as THE core asset and putting in place the platforms needed to collect, analyze, store, and use large volumes of data, audit firms will be able to reinvent the audit methodology and build a lasting competitive advantage for the future.
Recent Comments