The debate on process automation and “robot armies” are beginning to flood 21th-century workplaces and tech-savvy Enterprise, but does this extra muscle come with a whiff of science fiction.
Let’s look at the fact.

The proposition is clear. 
Instead of spending the working day on repetitive manual office work, finance professionals and knowledge workers can focus on helping their business, partnering to growth, improve and control. In theory, at least.
RPA on the rise…
RPA is all the rage right now. Short for Robotic Process Automation, it is an umbrella term for technologies to automate specific tasks with artificial intelligence.
Think of it as a personal agent, an app silently working for you while you work for the business. In its original form, it’s not an integrated part of the current or legacy software, but an interface that sits on top of the user interface or desktop, in applications like accounting and planning software, and acts just like the user “pushing” keys and executing tasks.
Easy, right? So who wouldn’t want to have a personal robot army?  Everyone who can afford it, and can afford the change management too and that is quite a cost.
1)    The implementation phase requires significant resources as robots need training and adjusting to improve process performance
2)    And robots are only as good as their input. Crap in, crap out and “computer says No!” type crap!
The deviation dilemma
Data quality and learning curves go hand in hand – for people and robots. In financial processes in general, and accounts payable in particular, deviation management is an unavoidable part of work. Typos, wrongly calibrated processes, or failing tech such as lousy quality scans creates all sorts of errors: missing PO numbers, wrong pricing, spelling, VAT confusion… the list goes on.
Deviations are a huge source of inefficiencies and a lot a threshold for automation. Hackett Group has reported that 0.017% of revenue is on average lost due to invoice error handling. In the EU, it equals to 237 billion euro, despite all kinds of automation tech. Moreover, if we add the Pareto principle, let’s say 20% of the transactions create 80% of the workload, we get a hint of the real cost in finance operations.
This is also why AI today isn’t a part of the core solution. You can apply as much smart tech as possible – it will do your organisation no good if the processes aren’t in order, and the data is of poor quality – meaning yours and to an extent your suppliers too.
Process control
We’ve previously discussed the importance of data quality. It’s not a problem if your current solutions can’t meet your requirements. With a completely digital exchange of transactional information, no critical information gets lost for future needs and requirements.
Your next step is to take control of systems and processes.
You might think you have a clear sense of how transactional data flows through your organisation; principles of data storage, formats, deviation analysis, sales and purchase data, workflows, security, but the truth is that most CFOs are uninformed on these process areas but they will hind process automation benefits.
So finance professionals need to take a hard look at their processes and understand the complex web of requirements, dependencies and technologies, that are central to improving and automating financial processes.
Step #3: Learn your processes

  • Map how transactional data is processed across your organisation
  • Chart your workflows
  • List systems and software
  • Identify complexities, bottlenecks, risks and costs

Get learning #4 here!
Co-authored by Christopher Argent, Generation CFO and Per Holmlund, Qvalia, a Stockholm-based fintech company that improves and automates financial processes. This month, Qvalia are launching Invoice Shield, a cloud-based firewall for transactions and the fastest solution for organisations to increase transactional data quality and reduce invoice errors, A/P workload, and legal risks.
More big questions here – #BQs