Part 2 in the Finance Automation series
Many companies are on digital transformation journeys because the future is certainly not analog nor based on paper. The finance function can’t escape a digital transformation journey either; for many, though, “digital” is still a buzzword rather than a reality. That’s why most CFOs will need to create a road map for digital transformation in their finance function.
Digital can be divided into three primary areas: the digital customer journey, new business models, and digital operations. Finance and accounting belong in the last category. Going digital is mainly driven by technology. However, processes and people are equally important. You need to find the right mix to create “finance masters.” In a previous article, we have broken down robotic process automation for you. It sparked a conversation around what technologies are available to transform finance, how they are different from one another, and when you should apply them on your transformation journey.
What’s available, and how can you use it?
Most of this is high-tech stuff and needs some explanation. What’s certain is that you need a structured approach, such as the following stepladder of available technologies based on complexity, when to apply them in your digital transformation journey, and what you can use them for.
One level is not necessarily a prerequisite for the other, so it’s likely that you can skip levels and jump up the ladder. However, this adds uncertainty to your transformation, as you cannot predict exactly how each of the technologies will fit into your existing process and systems landscape. To make it more tangible, here’s an example of how you could apply each of the technologies in your finance function.
Traditional automation starts within the ERP system, using some embedded process automation capabilities, which can be set up via customization or configuration. Within the ERP domain, you have several IT enablers created for specific purposes – for example, job scheduling solutions such as Control M, Automic, or SAP Business Process Automation by Redwood. They do nothing more than running jobs in the background based on job chains with dependencies.
A step further are IT enablers that provide rule engines that do a deeper analysis based on standard ERP functionalities. A good example is BlackLine’s Smart Close solution, which resides in SAP ERP and performs a rule-based analysis of output generated by ERP transactions. Furthermore, outside of the ERP domain are cloud-based solutions like BlackLine or TrinTech that serve a specific purpose such as account reconciliations. With those solutions, you can also reach high levels of traditional automation.
Robotic process automation is the level of automation that does tasks in the same way that a person would perform them. It mainly does rules-based tasks with a Boolean outcome: either right or wrong. Examples of RPA cases are comparing ERP data with data on the Internet such as bank reconciliations, T&E reconciliations, and many more. A popular one is for automatically creating timing or accounts payable accruals where you run ERP reports and download the information to an Excel spreadsheet, perform some calculations, and input this data into an ERP journal entry and post it. The best-known solutions are BluePrism, UiPath, Automation Anywhere, and Redwood Robotics.
Cognitive automation is the next step after RPA. In this phase of automation, you are transferring cognitive functions from a person to a system or a robot. You train the system or robot in a more advanced way to process and respond to information as a person would. Good examples of cognitive automation are the classification of incoming emails. Based on the classification, you then hand tasks over to RPA solutions to fully automate the solution using natural language processing engines.
Examples of solutions that provide this are the Cogito Intelligence Platform, IBM Watson, and ELIS. Furthermore, the use of chatbots is increasing. They are mainly used in customer service environments or in procurement to answer simple questions from customers or vendors. They are also used in accounting, but mainly for knowledge management. In this domain, a lot of suppliers are providing chatbots as a service. One of the main challenges in cognitive automation is the availability of good-quality data needed to train the system/bot.
Artificial intelligence is an interesting option. We are still looking for use cases of applications for finance and accounting where true artificial intelligence is used beyond cognitive automation. True artificial intelligence uses cognitive automation to train a system/bot and then, based on self-learning capabilities, the system/bot adapts its model to changing conditions without people’s interference. WorkFusion is one provider known for these capabilities with its Smart Process Automation Solution. Many more providers claim that they offer these features. Our opinion is that we still need to wait two to five years before AI becomes mainstream in finance and accounting.
Singularity, a concept that exists only in theory, describes when systems/bots transcend AI to the extent that they automatically adjust their models in such a way that people no longer comprehend at a pace people cannot keep up with.
We hope this provides a more concrete explanation of what a digital transformation could mean for your finance function. Are you ready to embark on a digital transformation journey of your finance function?
Are you already applying some of these technologies or looking to get started? We would like to hear about your use cases or answer any questions you might have to get started on the journey. Feel free to initiate the discussion below.