We are certainly on the cusp of a fourth industrial revolution which will change the way in which we all work. Whilst with the past three industrial revolutions, technology changes have been disruptive, they have not been net destroyers of job roles. Does Artificial Intelligence promise this?
The world of business intelligence (BI) is fundamentally all about providing the right information to the right user and at the right time.
Throughout my career as a Finance Director, CFO and accountant within practice, I have worked with hundreds if not thousands of SMEs and the one common theme is them not clearly understanding their business performance…..so is there now finally the opportunity to solve this?
The benefits to SMEs of receiving business information are now beginning to become clear with a far increased adoption and awareness of the power of their own business data.
We at Cascade.bi have proudly developed our out of the box dashboard platform which connects to Sage 50, and provides SMEs with easy access to their data, increasing awareness of their business metrics. The access to Sage 50 business intelligence dashboard vastly improves the decision making process.
(screenshot from the Cascade.bi platform for Sage 50)
This data could take the form of very simple Google Analytics, through to the number of Twitter followers or customer or product sales. Moving forward the understanding of these can be interpreted using AI, however I believe the limiting factor is that all businesses operate differently and the owners do not have the time to teach AI the basic tasks, or perhaps do not even understand the basic tasks to a sufficient level of detail themselves.
AI is undoubtedly a powerful tool which can very easily take on tasks within a business. I think it will be adopted mostly in its most basic form, through machine learning and the standardisation of repetitive and rules based tasks. These tasks for an SME are often performed by the finance function and therefore areas such as credit control and some reporting will be eroded by a far more efficient process.
The adoption of AI for many SMEs may be problematic in the short to medium term as they fail to understand the power it can bring and generally lack the trust of a “computer system”. The most sophisticated businesses will embrace the technology first.
Let’s first take the example of a simple credit control team within an SME. They will periodically raise and send statements, (checking them first for errors), email and phone the customers to chase for payment and use a strong element of human interaction, emotional intelligence and discretion to ensure the cash is received in a timely manner. Using AI to replace this whole function would eliminate the relationship and rapport element to credit control and may end up damaging relationships with customers. Instead a hybrid solution of utilising AI would work, providing prompts to action, automated emailing of documents, leaving the emotional intelligence and relationship aspects to the human. If trained properly AI could free up the mundane tasks and make the credit controller more efficient.
AI is currently in its early stages and is good at the simpler, limited problems like sentiment analysis (e.g. is a free text sentence from your customers good, bad or neutral) where there are few variables. This ‘opinion mining’ may not always be present in communications with customers with an increasing level of email communication being done with very limited amounts of verbose.
Non AI statistical methods are more than sufficient at stating where a business currently is (KPIs % increases, ratios etc.) but are unable to explain why the current situation has arisen or forecast the future.
AI has promising potential to explain both why and predict the future but current AI technology (and this is likely to be the case for at least the next 5 years) requires a lot of processing power (that even the explosion of cloud computing can’t help bring down to the real-time realm that BI requires) and a lot of data. This processing power coupled with the disparate data (paper based), will be a real obstacle.
AI may start to have application in larger companies with sophisticated systems that produce and store large quantities of data for AI to sift through but SMEs don’t. Their systems are usually limited and only tell a small part of the story, a lot still keep important records in paper form (outside AI’s reach), they rarely have specific workflow tools so their accounts system is the biggest source of SME data which will contain a wealth of information about who bought what product and when but without more (non-transactional) data about the purchase, few insights can be gleaned that standard statistical models cannot already predict.
An example for an SME, is that AI isn’t going to predict that sales from a customer will dip next month because the sales rep is off sick or that the contract to supply is due for renewal (held outside any system) and that the contract terms change, alter or the customer terminates their supply agreement.
I firmly believe that AI in its simplest form, can be used to assist SMEs in certain areas, however these are generally the more task based approaches, which would replace a human repetitive task.
In summary, whilst the next industrial revolution is upon us, the number of net job roles to be destroyed will take time. I think in 5 years time, SMEs will start to see the benefits, and it will be those leaner, fitter businesses with more technology accepting minds who will embrace it first and reap the rewards. The masses of “typical SMEs” will be slow to adopt and this will protect jobs for the next decade.