By Magali Michel, Director, Yooz
These days, automation is an indispensable tool for businesses looking to increase productivity and agility in a highly competitive economic environment. Gartner predicts that by 2020, robotics and automation will have become mainstream in finance departments, while AI is already having a profound effect for things like data mining and analysis.
The ROI is evident: cost reductions, optimisation of processes, data security, regulatory compliance,and much more.
That’s all very well, but, out of all the existing solutions, which one should you choose? How can you be sure you’re choosing the tool best suited to your needs?
Here are five all too common mistakes to avoid which set up finance projects to fail right from the start:
1.You use multiple software solutions for multiple projects
“For each task, the right tool”, said the Vietnamese philosopher Kaslan’tien. This was a long time ago, however, and this statement is no longer pertinent – at least in a business setting.Single apps can host several different tools under the surface, so the use of multiple solutions for different projects is only going to reduce efficiency through the time and cost lost to staff training on each different solution. Besides, it’s not like orders, purchases, expense reports and invoices have anything in common!
- You chose a solution which does not communicate natively with your accounting software or your ERP
As any standard IT department will tell you, integrating solutions into the existing data system is a nightmare, and it’s always preferable to avoid going down that route. Having a collection of different APIs, connectors, interfaces, micro-services and configurable modules creates a minefield of confusion for everybody involved. Make sure you have the right tools set up and that they all talk to each other before you make a final decision.
- You chose a solution which does not automate the invoice input and validation processes.
When it comes to data security and quality, businesses can’t afford to mess around. Security has become one of the biggest external threats to businesses of all size today. So why leave invoice entry and validation up to human intervention?
We all make mistakes, but data systems have proven that these manual, labour-intensive tasks are often better left in the hands of AI and machine learning systems that have been designed to learn and even pick up on mistakes made.
And, for those fearing that the robots are taking over, automated solutions have actually been predicted to create more than 58 million new jobs, meaning accountants may be tasked with more client-facing and customer relationship roles that can’t be handled by machines.
- You thought invoices couldn’t be processed on a smartphone
Smartphones are polluted enough already with games, apps and social media, so why would we want to add strategic work applications? “What happens in the company, stays in the company”, went the universal principle – if only out of concern for security.
But having a mobile workforce is becoming central to how a lot of business operate.
Instead of travel delays holding up invoice processing, managers could log into a mobile app and approve invoices before they even stepped into the building. Did the solution flag a suspicious email or account? This could be dealt with on the commute home rather than waiting until the next day.
- You avoid solutions which provide performance indicators
Developing a performance indicator is sometimes viewed as a painful exercise for finance departments. Firstly, for the people responsible for designing them, who suffer in order to collect the relevant data, but also for the people who sent them but never have time to read them.
Worst of all, what if the indicators continue to show that there is no continuous improvement and your teams are lacking in effectiveness? It’s enough to give you guaranteed chills.Everyone knows, that burying your head in the sand is not the solution to fix things. Performance indicators aren’t just there to show how badly things are going – it’s there to give an idea of what processes need to be changed and how. Having presentable, performance-led reports is a must if you are aiming for improvements.
5 steps to AI success
Each of these things are critical to projects running smoothly, and each has to work together in order for businesses to realise the optimal level of efficiencies from an AI-powered, automated solution.
If you’re reading this and one or even all of these points resonate with you, it might be time to go back to the drawing board!