Written by Gopalakrishna Tadiparthi, SVP of machine learning and artificial intelligence
In the rapidly evolving landscape of the CPG industry, AI is poised to revolutionize various aspects of the industry with the multifaceted advances in the technology. In my many years working in the field of AI and ML, it’s been true that CPGs are slow to adopt the technology for their analytics needs. But why is this when predictive analytics can offer so much to their business?
According to a 2023 RIS News survey, only 13% of the consumer goods companies surveyed felt they were adopting AI/ML technology to improve analytics, compared to 23% of retailers. The study specifically added that more than a third of brands using AI have not used the technology at all for pricing, demand planning and key functions like optimizing promotions.
[Source: 2023 RIS News Consumer Goods Technology Analytics study]
While there may be several reasons why CPGs haven’t fully invested in AI, one plausible reason is they haven’t armed themselves with enough knowledge and education before committing. It’s one thing to understand that AI will undoubtedly help a brand optimize assortments in stores, but it’s another to embrace and trust how AI will work inside a company’s existing infrastructure and processes (own plumbing).
From my view, as the head of AI/ML technology at Insite AI for three years — or lead plumber, if you will — and having spent more than 12 years at dunnhumby working on advanced analytics, I have three questions that all brands should ask an AI partner before working together.
These key questions will help brands better understand AI and implement a solution with more confidence.
Question 1: “Can you build the solution internally?”
At Insite AI, the answer is yes. One myth around AI is that a consumer goods manufacturer always needs to turn its data over to a source outside the company to receive AI-powered insights. They don’t. A dependable AI platform brings the toolkit to a customer’s house and integrates directly into their cloud. There’s no need to take the data out of the brand’s internal systems. Insite AI brings its AI expertise to CPGs and customizes algorithms that fit their architecture.
Many vendors and AI companies require that CPG send the data externally and require the CPG company to work on the standard solution. A major problem here is that there is no guarantee that a brand’s data will play nice inside that AI company’s solution.
Question 2: “Can you harmonize multiple, varying sets of data?”
For Insite AI, that’s affirmative. CPGs have rich data. They possess data from disparate sources and as a result the data. Converting heterogeneous to homogeneous data is a major benefit of AI done right. Predictive analytics and retail-focused large language models (LLMs) can take a wide range of data sources and harmonize the results all in one place to help tell a more succinct and effective story.
“At the heart of data is storytelling, and brands can struggle to find a story to tell if data is lying in piles around the house.”Gopalakrishna Tadiparthi, SVP of Machine Learning and Artificial Intelligence, Insite AI
An AI platform harmonizes this data all in one location. Just as Insite AI can come to a customer’s house to build a custom product that works, we also clean up the house, too. We bring our toolkit and pick up the scattered data, knowing which sources of information are best to use for specific problems and implement accordingly.
It’s paramount to have one AI platform distill varying data sources such as sales data, macroeconomic insights, third-party panel data analysis, weather, etc., into one location, one single source of truth.
Question 3: “Can your AI platform explain the data?”
Absolutely, with Insite AI, it can. After harmonizing the data, Insite AI’s proprietary AI model can explain the decisions. It’s not enough for any data or insights program to deliver results, the AI needs to explain the “why” behind the numbers. Insite AI prides itself on the models we use for “explainability.”
Traditionally, CPGs use statistical linear regression models. Through these models, they are telling stakeholders a story. But, with disparate data sets, it can be hard to tell a story. Insite AI has developed a three-step process, where we take all the data in, find what’s useful, and then build a machine learning model that makes the data parsimonious, or explainable for the end user.
How can CPGs lead in AI adoption
In conclusion, the pathway for Consumer Packaged Goods (CPGs) to excel in AI adoption lies in the judicious handling of their vast data resources. The integration of AI/ML and advanced predictive analytics transforms this data into enduring strategies for brand growth. Notably, an effective AI model must not only optimize data utilization but also prioritize efficiency and security.
CPGs should embrace change, select AI partners wisely, and collaborate with retailers to unlock shared value. Ultimately, the correct implementation of AI/ML not only enhances sales but propels brands toward the achievement of their annual business objectives. The convergence of strategic foresight and technological prowess positions CPGs to thrive in the evolving landscape of artificial intelligence.
Contact us to learn more.