Get your company ready for artificial intelligence

Companies need to resolve issues regarding data quality, ERP and business process improvement first, before implementing artificial intelligence solutions.

dominoes topple fall
Pixabay (Creative Commons BY or BY-SA)

Your next step is simple. You are the first domino. – Gary W. Keller

It looks like artificial intelligence (AI) has outranked all other technologies in popularity in 2016. It was a year where a broad audience became aware of its potential and risk. There are thought leaders who compare AI to innovations like electricity and the internet. The thinking is that AI is augmenting people in executing tasks. With that in mind, how does a company get ready for artificial intelligence?

AI has entered our work and personal lives in different formats and pace. For a company it is important to understand that AI is an "enhancer" that is dependent on other elements. Compare it to a domino stone track. AI is one of the stones, or perhaps many of the stones, but it needs other stones to fall first.

Companies struggle to understand what those stones are and in what order they need to be placed? Or even worse, there can be parallel tracks of domino stones that rally at the same time and draw on the same pool of resources. Yet AI continues to evolve rapidly and a situation of "being stuck in the middle" puts industry peers ahead of your company. What should you do?

The first step is to understand the common denominator of all AI technologies. They all rely on massive amounts of data. The good news is that companies collect data at a rapid pace and in amounts that are growing year over year. The bad news is that the quality of data is not meeting the needs of most AI technologies. Accuracy, completeness, relevance, consistency, reliability and accessibility are aspects of data that are a huge challenge for any company. 

You see that some companies appoint a Chief Data Officer to deal with "the data problem." That is a good step forward, however it is not fixing the root cause. If you peel the onion, you will find a number of reasons that pollute quality of data: poor data definitions, inconsistent and sub-optimized business processes, and under-utilized core business applications like enterprise resource planning (ERP) are the prominent ones.

Data quality is one of the domino stone tracks that a company has to set up and rally, but there are more. At a certain point in time the data track has to collide with the "operating model" track. What do I mean with that? Companies have to answer the question if they have to be a "real-time business." AI technologies demand your business to be "real-time."

A "real-time" business is able to perform transactions and run analytics at the same time. As soon as you have entered a sales order for a certain product or service, the company is also able to understand sales performance. Or another example, for a manufacturer the real-time data collected from sensors in the production process, enable the production planner to take corrective action when needed. Predictive analytics are based on AI technologies and are core to these kind of solutions.

However these predictive analytics solutions would not be meaningful if the core ERP system is not able to perform "real-time" transactions. This is the second domino stone track that the company needs to rally: upgrade or re-implement your core ERP systems. The major ERP software providers have recognized this business need and have re-build their solution offerings. Their ERP solutions process transaction in-memory which means that transactional data is instantly accessible for analytics and subsequent transactions.

So now we have two domino stone tracks running in parallel: data quality and ERP. There is another track that the company has to initiate to ensure that there is a solid base-line for AI solutions: consistent use of a simple, standardized and harmonized set of business processes.

This third track is as challenging as the other ones. Companies struggle to standardize and harmonize business processes. Silo-ed behavior of functional disciplines, lines of businesses, and operating units is the main reason. Cross-functional integration is a prerequisite to optimize business processes to a world-class level, however many companies are still optimizing their verticals. A cultural change has to happen first.

All of the three domino stone tracks, data, ERP and business process, need to rally first before any AI solution can be effectively deployed. Make sure that you are doing this groundwork first. AI pilot projects can rally as a parallel domino stone track to keep the company engaged on a fascinating emerging technology. But foremost the focus must be on getting the house in order first.

This article is published as part of the IDG Contributor Network. Want to Join?

SUBSCRIBE! Get the best of CIO delivered to your email inbox.