AI for CPG R&D Innovation: 8 Considerations in the Build-vs-Buy Conversation
In today's ever-evolving business landscape, digital transformation initiatives and harnessing the power of AI have become increasingly important for companies in the consumer packaged goods (CPG) industry. It's worth noting that according to McKinsey, around 70% of digital transformation initiatives tend to face challenges and fall short of their intended goals. Similarly, Gartner suggests that approximately 85% of AI projects also encounter difficulties or fail to meet expectations. If done right, AI can help improve sales, optimize operations, reduce costs, and empower organizations to create new products and even pursue new markets.
Understanding these statistics is crucial because a successful AI project often relies on the organization's investment in data and technology. The combination of robust data and AI capabilities can be a game-changer for CPG research and development (R&D) teams, enabling them to improve operational efficiencies, accelerate time to market, reduce costs, enhance customer satisfaction, and gain a competitive edge over their rivals.
While some internal IT and data science teams will try to build systems and potential models tailored to the specific business requirements, others may choose to leverage the expertise of external vendors. Each organization involved in this process must thoroughly explore and evaluate the advantages and disadvantages of both building and buying to make an informed decision.
At Turing Labs, we have worked with some of the top CPG companies in the world, and here are some of the lessons we learned to help provide some guidance on things to consider when buying an AI solution or deciding to build one in-house.
1. Total cost of ownership
When deciding between building your own AI-based software or purchasing a commercial solution, it is crucial to consider the total cost of ownership. This factor is often underestimated in the case of building non-SaaS applications. Custom development entails significant upfront expenses, including system design, implementation, and the recruitment of specialized talent may require engaging with high-end consulting firms. This is especially true for AI-based software, as it demands resources from skilled data scientists, ML engineers, dedicated development teams, and IT infrastructure that may not be readily available. On the other hand, commercial AI solutions offer more predictable and scalable cost structures. Vendors often provide tiered pricing models that can adapt to the organization's evolving needs, allowing for a more flexible cost management approach. This reduces the reliance on internal resources required for in-house development.
2. Innovation is a continuous journey
Another thing to consider is that commercial AI solutions enable proactive innovation by leveraging vendor expertise and continuous upgrades with cutting-edge technologies. In contrast, internal innovation may lag due to resource limitations and organizational constraints. Implementing a commercial solution is typically faster, leading to quicker ROI realization, while building custom applications can result in longer development cycles and potential goal misalignment. Moreover, the need for continuous maintenance and system upgrades to keep pace with AI advancements adds to the costs of in-house development. While commercial solutions provide continuous support, updates, and technical assistance, ensuring access to the latest features and innovations without relying solely on internal resources.
3. Focus on core business
Let’s remember CPG companies are typically not software companies, so building custom AI applications can divert resources and attention from their core business of creating market-leading consumer products. This may lead to a loss of focus on primary business objectives and differentiators, which is particularly important in the competitive environment of CPGs. Additionally, the process of building AI software requires technical expertise in AI technology development, machine learning techniques, and model development and testing. It can be challenging and time-consuming for teams not specialized in AI. Moreover, diverting teams from their core focus, such as IT and R&D, may impact other aspects of the business.
4. Ecosystem of innovation
When evaluating whether to build your own AI-based software or opt for a commercial solution, it is important to consider the ecosystem of innovation. Choosing a commercial AI software vendor allows for collaboration and co-innovation while respecting the intellectual property (IP) of both the company and its vendors, such as suppliers. Commercial vendors often have extensive networks and partnerships, enabling them to play a central role in driving innovation within the ecosystem. This collaborative ecosystem of vendors, partners, and customers, encourages the exchange of ideas, insights, and expertise, fostering a mutually beneficial environment for all parties involved. In contrast, building your own software may limit the potential for co-innovation within a broader ecosystem.
5. Cost of learning
If your organization has the capacity to continuously learn from industry best practices and keep up with leading-edge developments, building your own software may be feasible. However, commercial software vendors specialize in AI technologies, investing substantial resources in research and development to ensure their software remains at the forefront of advancements. For organizations whose core competencies lie outside AI development, matching this pace can be challenging. Additionally, commercial vendors bring extensive expertise, having resolved diverse problems for various clients. This accumulated knowledge often exceeds that of an internal team focused on a single project. Furthermore, commercial AI applications offer the advantage of industry best practices, refined through experiences with multiple clients. This enables swift identification of trends and iterative improvements, benefiting all users. Conversely, building a custom AI application restricts an organization to insights derived solely from its own experiences.
6. Leveraging connected systems
Building a connected system with a clear product goal in mind is highly valuable compared to developing IT tools that may not align with the language and needs of the business and R&D teams. A connected system focuses on integrating various components and functionalities, enabling seamless communication and collaboration between different stakeholders. This ensures that the software aligns with the specific goals and requirements of the business and in many of the cases we deal with, the CPG R&D teams. What this does is allow them to effectively leverage AI technology to drive innovation and achieve desired outcomes. This can oftentimes be difficult in large organizations when determining ownership, control, and budgets. Trying to build specific tools for specific functions is oftentimes under-delivered in these larger organizations, due to disconnected IT tools that lack business and R&D language can hinder communication and limit the overall effectiveness of AI-based software.
7. Understanding AI maturity
It is crucial to consider your organization's AI maturity. Understanding where you currently stand in terms of AI capabilities and experience is essential. Many companies have learned from past experiences and have transitioned from a build-first approach to a buy-first strategy. The primary driver for this shift is the rapid pace at which AI technology evolves and the need to keep up with future advancements. By opting for ready-to-use domain software, companies can gain a competitive edge by quickly scaling their AI initiatives ahead of their competitors. This strategy allows organizations to leverage existing expertise and proven solutions, saving time and resources while staying at the forefront of AI innovation.
8. Leverage R&D formulators as bridge builders
To rapidly respond to the changing consumer and CPG market, it is essential to equipping product developers(formulators), with robust tools capable of delving into complex data objects. By utilizing AI-based software, these formulators gain access to a wealth of hidden information, enabling them to refine their hypotheses and theories regarding the factors that influence system behavior. Moreover, such software provides a collaborative framework, fostering seamless integration with other disciplines. Through intelligent data integration, it is important for commercial AI systems to be able to combine human expertise with the vast knowledge encapsulated within the organization. This captures critical subject matter expertise and tribal knowledge, facilitating informed decision-making in both the short and long term. It is particularly valuable because certain aspects of the decision-making process are inherently ambiguous and challenging to model or automate. By opting for AI-based software, companies can effectively bridge the gap between data-driven insights and human intuition, thereby gaining a competitive edge in the dynamic marketplace.
In conclusion, when considering whether to build your own AI-based software or purchase a commercial solution, CPG organizations must carefully consider these 8 key factors. Harnessing the power of AI can provide a significant competitive advantage, but it requires a thoughtful approach. By examining these considerations and learning from the successes of industry leaders who have embraced commercial AI solutions, CPG companies can make informed decisions that drive growth, improve customer satisfaction, and outperform their competition in the dynamic landscape of the CPG industry.
About the author(s).
Manmit Shrimali
Co-Founder, Turing Labs Inc.