In the beginning of the year Snowfox participated in a study through a master’s thesis, that investigated experiences and perceptions on AI in the financial administration domain. Although we at Snowfox have a reasonable understanding on the experiences of financial administration personnel regarding the subject, an academic study gives a much-needed perspective into the new winds blowing in the financial administration field.
In the now-beginning blog series we look at the findings of the study from three key point of views and especially how financial administration personnel see the future of their domain. The study in question can be obtained from the database of Jyväskylä university’s library.
It should not come as a surprise to anyone anymore that it is possible to gain massive savings and a significant increase in efficiency in the financial administration domain through the use of AI. On the other hand, AI has not yet solidified it’s presence in financial administration processes and due to that it is often seen as a incoherent or even scary new component.
By far the greatest challenge in utilizing AI in financial administration seems to be the lack of knowledge on the subject. It seems that financial administration organizations often are not able to desire AI-based solutions, simply because there is no sufficient know-how on the subject within the organization.
AI also differs from other automation, as it makes it possible to solve new types of, more abstract problems automatically. Perceiving this can be challenging if one looks at AI from the point of view of traditional automation. When the potential of the technology is not recognized, essential or at the least sensible development investments are postponed, leading to challenges in terms of staying on par with technological development.
When the lack of know-how is mixed with the disorderly public discussion on the subject, it’s no wonder that non-experts often struggle to catch the true essence of AI. Academics often define artificial intelligence as a disruptive technology, regardless of the domain it is used in. As the name suggests, these kinds of technologies often disturb the domains old, “normal” way of doing things, which leads to a rapid pressure for change and almost always to a need to adapt into the changing environment.
If using AI is challenging and understanding it is hard, why should it be used in the first place? Why add challenges to current processes that are already working well? In addition to everything presented above, especially due to two reasons. Usually, the existing challenges are a product of processes not working as well as organizations tend to think and because using AI will be inevitable in the future for upholding efficient financial administration processes.
The argument of poor processes within financial administration organizations sounds unfair, but it can easily be demonstrated with a simple example that came up in the research countless times. Before implementing any AI solutions to financial administration processes, when you ask from a purchase ledger team with what the percentage they can correctly post invoices going through them, the answer is usually something very close to one hundred percent. When the same process is then tested with AI-assistance, in many cases it is quickly discovered that the success rate of manual labor is often not even close to what was believed. In many cases the AI-based solution has even been more accurate than manual labor. The phenomenon is amplified if business functions are responsible of posting purchase invoices.
Based on the study, especially the managers of financial administration organizations have a clear consensus of the direction where the domain is heading. Using AI is seen as a necessity for providing cost-efficient and high-quality financial administration services in the future. Surprisingly, it in fact seems that excessive caution with new possibilities is a greater risk than building the capacity only after the technology has solidified its place within the domain.
In the next part of the blog series, we take a look on the challenges of utilizing AI and how they can be solved. Spoiler alert, the challenges are relatively small and easily solvable.
The same way as in all the other, similar challenges. Many financial administration organizations also have no own legal or IT-department, as they are rather outsourced from an external partner based on changing needs. Creating a working AI-solution to a specific problem per se requires so much resources and personnel with specialized skill sets, that even for a large organization it is usually more sensible to acquire a proven concept from the markets, rather than developing a solution internally.
When choosing the partner for the project, it’s important to think about the big picture. Is the offered solution long-lasting? Will it be further developed in the future? Does it create genuine added value from day one? Can you test the solution with your own data in your own environment before making the decision? Can you get out of the contract easily if need be?
The question shouldn’t anymore be if AI is needed in financial administration, we should rather be considering how long we can go without. It’s worth to start utilizing the potential of AI right now, in a way that fits the needs of your organization, so that the force of change doesn’t slam on your face all at once in the future. A good first step for beginning the AI journey is to consider the possibilities of the technology in the automatization of posting and routing of purchase invoices, where it’s possible to gain major advantages with a low financial stake, at the same time preparing your organization for the future.
The Snowfox.AI service can route and post your purchase invoices automatically with artificial intelligence. You no longer have to worry about manual tasks.