Julia Chirkova
The seemingly innocuous phrase "I just pull the figures out of Xero into Google Sheets" often belies a significant amount of manual labor within accounting departments. For finance professionals, this simple description can mask a complex and time-consuming routine that, when meticulously mapped, reveals a substantial drain on efficiency and valuable human capital. This is precisely the challenge faced by the finance team at Latch, a company that has successfully leveraged Artificial Intelligence (AI) to transform a once-tedious 15-minute process, requiring over 200 individual clicks, into a streamlined operation demanding a mere 5-10 clicks. This transformation serves as a powerful case study for how AI, when strategically applied, can address pervasive issues within the accounting profession, particularly in the face of a growing talent deficit.
The core of Latch’s efficiency breakthrough lies in its strategic application of AI not to replace existing software, but to act as an intelligent intermediary. For years, the process of updating financial models involved a laborious sequence of actions: setting date ranges, exporting data in CSV format, meticulously deleting unnecessary periods, writing complex lookup formulas, and wrapping values in error checks – all performed manually, week after week. While each individual step might appear minor, their cumulative effect created a fragile, repetitive chain of manual effort. By implementing AI to bridge the gap between their accounting software (Xero) and their spreadsheet platform (Google Sheets), Latch was able to automate the execution of these individual steps, collapsing the extensive click count and significantly reducing the time and cognitive load associated with the task.
This experience highlights a critical realization for many finance leaders: while the transformative power of AI in accounting is widely acknowledged, identifying a practical starting point can feel overwhelming. The narrative from Latch suggests that a complete overhaul of existing software stacks or the implementation of complex, enterprise-wide platforms are not prerequisites for achieving immediate, tangible results. Instead, the most effective approach often begins by scrutinizing and optimizing the processes already embedded within the current financial close cycle. By providing AI with a clear understanding of their specific workflows and the context behind each action, Latch was able to identify and eliminate organizational pains, thereby eradicating the rote, repetitive tasks that historically burdened their close cycle.
Bridging the Talent Deficit with Existing Systems: A Growing Imperative
The need to find practical solutions for eliminating manual friction in finance departments is not merely a matter of convenience; it is rapidly becoming a mandatory requirement for sustained growth. The accounting industry is grappling with a significant talent shortage, a reality underscored by recent industry reports. According to the AICPA & CIMA’s 2025 Trends Report, a notable decline in accounting degrees awarded by higher education institutions has exacerbated an existing supply-demand imbalance. In the 2023-24 academic year, 55,152 accounting bachelor’s and master’s degrees were conferred, representing a 6.6% decrease from the preceding year. Even if these numbers stabilize, the fundamental gap remains substantial: the accounting profession is expanding at a rate approximately seven times faster than the undergraduate student population from which it draws new talent.
This widening chasm means that finance departments can no longer rely on simply adding more headcount to manage increasing transaction volumes and the complexity of modern financial reporting. Efficiency must now be driven by intelligent system design and the strategic adoption of technology. Fortunately, the solution often lies not in acquiring entirely new, cutting-edge software, but in optimizing the capabilities of existing tools. Applications such as Xero, QuickBooks, and Excel are inherently robust in handling core accounting logic. The primary source of operational drag, therefore, is not the inadequacy of these individual systems, but the manual "empty space" that exists between them. This space is characterized by the repetitive tasks of copying, pasting, and reformatting data – the essential but time-consuming work required to enable disparate systems to communicate effectively.
AI emerges as an ideal solution to act as this crucial "connective tissue." It possesses the ability to extract, format, and route data seamlessly across existing software applications, automating processes that were once manual and prone to human error. Importantly, this technological integration does not diminish the role of human accountants. In Latch’s model, human oversight remains paramount; accountants continue to review the numbers, evaluate variances, and ultimately approve the financial close. The technology serves to remove the mechanical, repetitive aspects of data preparation, freeing up valuable cognitive resources for higher-value analytical tasks.
Five Questions to Ask Before Automating Financial Processes
Before embarking on any automation initiative within a finance department, it is crucial to move beyond superficial checklists and delve into the specific decisions and nuances that underpin each manual action. By asking targeted questions, finance professionals can accurately assess a process’s readiness for automation and identify the most impactful areas for intervention. The following five questions provide a framework for this evaluation:
-
What is the precise trigger for this task? Understanding the exact event or condition that initiates a particular workflow is fundamental. Is it a specific date, the receipt of a document, or a change in data within another system? A clear trigger ensures that automation is initiated at the correct moment.
-
What are the explicit rules and decision points involved? Every manual process, no matter how simple it appears, often involves a series of implicit or explicit rules. For instance, "If the invoice amount is over $1,000, flag it for manager approval." Identifying these rules allows for their direct translation into automated logic.
-
What data inputs are required, and where do they originate? Pinpointing the exact data fields and their source systems is essential for AI to effectively gather and process information. This involves understanding the structure and format of the data from each source.
-
What is the desired output, and in what format? Defining the exact structure and format of the final output is critical. This includes specifying the destination system, the required fields, and any necessary transformations or aggregations.
-
How is accuracy verified, and what are the common error types? Understanding the current error detection and correction mechanisms is vital for building robust automated processes. Identifying common errors allows for the implementation of appropriate validation checks within the automated workflow.
By systematically answering these questions, finance teams can gain a granular understanding of their existing processes, enabling them to design and implement AI-driven automation solutions that are both effective and precisely tailored to their needs.
Starting Small to Build Momentum: The Power of Incremental Automation
The ambition to "transform finance" can often lead to broad mandates that falter under the weight of their own scope. Attempting to solve too many problems simultaneously can lead to inertia and a lack of tangible progress. In contrast, focusing on a single, recurring process offers a far more conducive environment for learning, iteration, and building momentum. The ideal candidate for initial automation is a tedious, rules-based workflow that is thoroughly understood by an experienced team member – a task they would gladly relinquish to technology.
The recommended approach involves meticulously recording the exact steps involved in this chosen workflow in real-time as the work is performed. This detailed documentation should include clearly defined review checkpoints, allowing for human oversight at critical junctures. The automated flow should then be rigorously tested using live historical data, with the results systematically compared against the established manual baseline. Success should be measured not only by the reduction in manual touches and error rates but also by the increased capacity reclaimed during peak periods, such as month-end or quarter-end close.
Reducing a single, 15-minute, 200-click routine to a 5-10 click operation might seem modest, but the cumulative effect of such improvements across an entire year can be substantial. It translates into hours of high-value focus that can be redirected towards more strategic activities, including forecasting, variance analysis, and providing operational support to other departments. These are the activities that truly drive business value and differentiate a finance function from a mere record-keeping department.
Broader Implications for the Accounting Profession
The successful application of AI in streamlining manual workflows, as demonstrated by Latch, carries significant implications for the broader accounting profession. It signals a shift away from labor-intensive data handling towards a more analytical and strategic role for finance professionals. As AI continues to mature and become more accessible, accounting firms and corporate finance departments will increasingly be expected to embrace these technologies to enhance efficiency, reduce errors, and improve the accuracy and timeliness of financial reporting.
The ability to automate repetitive tasks also has the potential to democratize access to sophisticated financial analysis. By reducing the burden of manual data manipulation, smaller businesses and startups, which may not have the resources for large finance teams, can leverage AI-powered tools to gain deeper insights into their financial performance and make more informed business decisions.
Furthermore, this trend could reshape the educational and professional development landscape for accountants. There will likely be a growing emphasis on skills such as data analytics, AI literacy, process optimization, and strategic thinking, alongside traditional accounting principles. Professional bodies and educational institutions will need to adapt their curricula and certification programs to equip future accountants with the competencies required to thrive in this evolving technological environment.
The narrative from Latch underscores a fundamental truth: every accounting process, when examined closely, reveals a level of mechanical complexity beneath the surface-level summary figures. The job of finance professionals is to understand these mechanics and to continuously seek ways to optimize them. By embracing a systematic approach, asking the right questions, and starting with the most repetitive routines, organizations can embark on a journey of automation that not only reduces click counts but also liberates valuable human talent for more impactful contributions. The future of accounting is not about replacing human expertise with machines, but about augmenting human capabilities with intelligent technology to achieve unprecedented levels of efficiency and strategic insight.
