Experiences using artificial intelligence to automate administrative work have demonstrated that in the best cases, it is possible to automate up to 94% of the tasks involved with processing of purchase invoices, for example. However, many companies still have a long way to go before they reach this goal, especially if they have not looked after the harmonisation of data at an early enough stage.
If purchase invoices include standardised product names, account specifications are handled consistently and purchasing processes are implemented, it is much easier to create an algorithm even from a small amount of data. However, this means that different functions (e.g. purchasing and finance) and business units must work together and understand how important harmonized and standardized data is for the company’s competitiveness.
At the same time, organizations must understand the importance of data ownership and availability, in particular, when some of the operations are outsourced.
Utilization of machine learning starts with harmonization of processes
Ten years ago, many Finnish companies announced that they are outsourcing their financial administrative work to emerging countries, in order to gain labour arbitrage. These decisions were often solely based on cost-efficiency and very seldom on targeting, cross functional process development.
At that time, nobody could imagine how these decisions would impact on the opportunities to utilize machine learning today. To my understanding, in most of the cases, the development of processes was not a top priority, neither within the company, which had outsourced nor in outsourced service centres. One could conclude that some of those Finnish companies have wasted almost ten years of development work when it comes to harmonisation and standardisation of financial operations.
One exception to this trend was Wärtsilä which centralised its global financial shared service centre in Vaasa in 2008. This bold decision was based on the belief that productivity can be achieved trough innovative process design, automation and co-operation with business units. Which all were indeed, to some extend, realized.
I was myself involved in making this happen, and I am confident, that Wärtsilä can collect such unique benefits today, that where unseen in 2008. These are, for example, cross business harmonized processes and data accuracy.
Fortunately, many of the companies that previously outsourced their operations are now transferring or are planning to transfer these tasks back to Finland, thanks to the productivity benefits gained from software robotics and machine learning. Based on this phenomena, it is valid to claim, that the utilisation of artificial intelligence will generate work to Finland, but the skills required are of course different.
Leadership capabilities are crucial while implementing artificial intelligence
This is, above all, a question of leadership. If the company’s top management and board does not understand how artificial intelligence can be utilized in generating new business opportunities, the operative organization cannot make it happen.
The utilisation of artificial intelligence requires a new way of thinking especially in the top management. Firstly, they need to understand how these novel technologies will impact on their businesses and what kind of new competitive advantages could be achieved trough utilization of AI. Secondly, they must ensure, that organization has all the means to recap new skills and competencies needed for this chance.
Leaders must have courage in setting enough ambitious targets, encourage employees in their learning process. Addition to this, leaders must ensure the availability of data and, finally, make sure that the entire organisation understands that objectives can be achieved only by working together, not by sitting in silos.
The writer is a member of the Artificial Intelligence Programme, a Doctor of Science (Tech.) and a Master of Science (Econ). Merja Fischer has been in a leading position both in Finland and abroad (ABB, Nokia and Wärtsilä). For the past 2.5 years, she has worked as Director of RPA and AI at Staria Oyj.