Note: Reprinted from CPG Matters (May 2017).
By E.J. Kenney and Don Gordon
Everyone talks about how Artificial Intelligence (AI) and machine learning have the potential to unlock significant value, but recently they have become industry buzzwords without a clear definition. Even though AI and machine learning are hot topics, consider why they are important. To understand the importance of these technologies, consumer product (CP) companies should revisit the “Why” and “How” they can use AI and machine learning.
First, remind me: What’s the difference between AI and Machine Learning?
AI is defined as the ability of computers to mimic human logic and thinking. Machine learning is a subset of AI that refers to the ability of computers to learn from data without being programmed. Machine learning algorithms adapt to change by learning continuously as data accumulates, making them a powerful tool for CP companies seeking to make accurate predictions or take the next best action automatically.
Early machine learning efforts were plagued by speed and accuracy issues. Recent technology advancements have made it possible for machine learning to manage the high volume, velocity and highly granular internal and external data CP companies have available to them today. Machine learning solutions can deliver analytical solutions with massive computing power to process this data in real-time and automatically spot patterns and exceptions at both a rate and a degree of accuracy well beyond what any human could deliver.
Why: It all comes down to data
Every year, the volume of world data doubles, and 75 percent of it is in unstructured formats such as text, voice, and video. CP companies have never held so much information on consumers, suppliers, and internal processes. The question is: How can they harness this flood of data to achieve better business outcomes and compete more effectively?
Future winners in the CP industry will be able to harness all their customer and consumer data to pursue new opportunities, as well as overcome complex challenges, such as:
- Extreme pressure on margins; growing consumer expectations; and demands for greater agility (both internally and in dealings with customers) across the value chain
- Digital disruption caused by smaller, specialized CP firms that are bypassing traditional distribution models in favor of straight-to-consumer, e-commerce-first strategies
- Growing consumer demand for personalized products and simple, consumer-centered experiences.
Companies that meet all these needs will build lasting consumer relationships. While many CP companies struggle to scale personalization across products and services, despite access to consumer data, machine learning can help innovate and win in this complex business environment.
How: Unlocking the true potential
The successful use of machine learning should be defined by the value provided to consumers. They want things done for them in a way that’s faster, easier, and more personalized – so the better CP companies are at providing this value, the more successful they will be.
For example, apps that advise consumers when to buy flights are perceived as high value because they free consumers from the searching-and-matching exercises that take time and effort. Applying this concept to the CP industry will drive innovation and define the leaders of the pack. Imagine getting a reminder to purchase diapers, along with an offer for those diapers, before realizing they are needed. Machine learning can monitor individual consumers’ consumption and re-order rates, learning optimal re-order points and proactively provide consumers with personalized notifications and offers.
Similarly, machine learning is disrupting marketing processes in positive ways. Machine learning is increasingly behind in-store apps that can tell how long a customer has been in a certain aisle, allowing the app to then provide targeted offers and recommendations based on data about their personal consumption and preferences. CP companies are using machine learning capabilities to do things like set shelf pricing, determine product assortment and mix, and optimize trade promotions.
In addition, apps that are tracking consumers’ movements in-store and monitoring wait times in checkout lines are providing invaluable insights to both CP companies and retailers about store traffic and merchandising effectiveness at the individual store level. This enables both parties to collaborate to deliver increasingly tailored assortments and store layouts to maximize basket size, satisfaction and sell through.
Let’s take the concept online. When a shopper goes to a website in search of a product, the site must be able to push out personalized content that is highly tailored. Machine learning can determine which products may be of most interest to that shopper by comparing their profile with that of other similar consumers. It can even organize reviews to show those most useful to a shopper based on their preferences, as well as price products at an optimal price point based on what it has learned about that shopper’s price sensitivity. This type of personalization cannot be scaled without machine learning algorithms and AI embedded within all aspects of the customer experience.
The time is now.
While machine learning and its use cases have been around for years, it is only recently that CP companies have been able to harness it. It would have been impossible for humans to leverage the benefits of machine learning without recent advancements, such as affordable massive computing power to process massive amounts of data quickly. In addition, the ability to embed sophisticated algorithms within transaction processing to make and automate decisions in real time is a recent advancement, as well as analytics that are more than 98 percent accurate at predicting outcomes when processing the right data.
Given these technological advancements, there has never been a better time for CP companies to adopt the technology. And, according to McKinsey, CP companies that are early adopters of machine learning have seen significant positive results, such as dramatically improved sales, profitability, customer engagement, and productivity. So, what are you waiting for?