Driving Preparedness and Enhancing Data with AI
As consumer goods companies invest in their tech capabilities to better prepare for the future, more are looking toward the potential of AI-powered solutions. From improved visibility into consumer demand, which can better inform product innovation, to more efficient manufacturing and marketing, AI can transform processes across the organization. But are CGs prepared?
While AI is being implemented industry-wide, there is confusion over how to best leverage the technology and how to ensure the organization has the necessary framework for supporting scalable tech. Conor Keane, president and CEO of Spring Global, shares best practices for implementing AI within the consumer goods space, and how the technology can help drive business and streamline processes.
CGT: What's driving the growing emphasis on artificial intelligence within the consumer goods space?
Keane: AI is now a mainstream technology, so of course, every industry is adopting it to enhance their productivity. CPG is no exception; in fact, CG may be better placed to adopt AI than some industries due to the nature of its business: big data sets, repeatable processes, and existing IT sophistication.
The impact of COVID has increased motivation to use AI. With labor challenges, rising costs of goods, supply chain issues, and an ever-changing retail landscape, we simply have to use AI to be more competitive. We almost have no choice in the matter because competitors, customers, and key channels already use it.
CGT: How can consumer goods companies more effectively navigate the growing influx of data to make better insight-driven decisions?
Keane: You need to speed up in gathering, interpreting, and acting on data. You need your data to be LIVE so that you can see what's developing as it happens. Critical data should be streaming in real-time; you should use AI to monitor that data and detect anomalies as they occur or anticipate them before they occur, analyze, and do root cause to take corrective action.
So, the steps are:
- Build the "data pipes" that gather the data. It must be from multiple sources if you are going to analyze a problem properly. (Sources: Field tools, Business systems, Warehouse/Production systems, Support systems, in addition to cross-referencing with current weather and traffic conditions).
- Build the anomaly detection algorithm
- Build the root cause analysis algorithm
- Build playbooks that are the set of actions available to solve a problem
- Execute the playbook
- Measure the outcome and learn for next time
Data gathering and cleaning is hard. It requires process and technical rigor that takes years to develop. Your AI is only as good as your supply of relevant and clean data. This requires a mindset change. It is different from BI, where static reports are suitable for the specific use-case they were designed for, but they are merely data from a single point in time.
CGT: How can CGs instill an understanding of AI across the enterprise?
Keane: AI is not some mythical thing that is hard to understand. People use AI in their day-to-day lives as consumers when using Amazon and Netflix. So start by educating your management layer by showing them how AI works at a very basic, first-principles level. Make tools and talent available to harness ideas — democratize AI in your company. Stress that AI is not owned by your data scientist or BI team.
Show people that AI can turn data into recommended actions and allow them to see how those actions play out in the field. Genuinely understanding the value of AI means knowing that it continues to learn and grow with your company with the help of your SMEs.
AI should enhance and fuel your company into a better tomorrow; AI is not about replacing people. It helps make people smarter. It might lessen the need for some skills but will create needs for other skills. Be upfront with this.
CGT: Where do partnerships fit in for CGs looking to adopt AI?
Keane: There is no need to think that you need to build AI models on your own — you wouldn't try to build your own car engine, would you? The power of a good partnership means you benefit by collaborating with those who have all the industry best practices. They have spent years working with the best brands to build AI models that accelerate growth and increase efficiency in the workplace. This is a perfect case of "work smarter, not harder."
Don't be bashful about benefiting from other companies' challenges and failures.
