Practical guidance for transformative GenAI in R&D

How can generative AI (GenAI) be harnessed in ways that go beyond incremental productivity gains to also tackle challenges once thought unsolvable?

This question is front of mind for innovation teams across many industries. And Sagentia’s CTO Forum has been exploring the challenge to identify practical answers.

Our Generative AI Playbook for R&D sets out a path for progressive, value-driven use of the technology, drawing on CTO Forum members’ real-world experiences. We’ve distilled their insights and learnings into ten principles that lay a path for progressive, value-driven use of GenAI in R&D applications.

Largescale, transformational use of GenAI demands both ambition and pragmatism. Download our playbook for guidance on how to strike the right balance throughout the adoption journey.

With thanks to CTO Forum members from Amcor plc, Bayer Crop Science, Mars, Incorporated, PepsiCo Inc., Stepan Company, and The Proctor & Gamble Company for sharing their GenAI experiences and expertise.  

You can download a PDF of the entire playbook here, for free: Sagentia_GenAI-in-RD-Playbook-digital-Oct25-v2.pdf

Below is an excerpt from the Playbook.

Using this Playbook

Many R&D leaders are asking how they should harness the power of Generative AI (GenAI) within their organization. This Playbook aims to embrace that challenge – drawing on insights and experiences shared by members of our CTO Forum. We begin with opening remarks which set the scene for insights that follow. The Playbook then has three sections or ‘Plays’, each with an overarching message for R&D leaders. Under each Play we present:

• Observations – reflections on the current situation

• Principles – key learnings and guidance

In addition, in Play 2 we include a set of use cases provided by Forum members. We conclude with closing remarks, where we briefly review the current landscape for GenAI in R&D and outline potential topics for future discussion

The Plays: 1, 2, and 3

Play 1: Fit with digital journey Relating GenAI to other digital tools and initiatives, with a focus on the major challenges associated with GenAI.

Play 2: R&D problems to solve Providing guidance on when and how to use GenAI.

Play 3: Successful implementation of GenAI Explaining what is different about GenAI’s adoption and suggesting approaches to help deliver successful outcomes.

Definitions

• Generative AI creates new content ranging from text, images, and software code through to research hypotheses, experimental (synthetic) data, and new product concepts by ‘learning’ from existing data sets. Agentic AI (AI that can perform tasks autonomously) is treated as an implementation of GenAI

• Analytic AI can process, analyze and interpret structured, statistical data – using algorithmic approaches such as machine learning, natural language processing and data mining to identify patterns, generate insights, and make predictions

Introduction

GenAI heralds a new era for R&D and innovation – with an expectation of far-reaching change across R&D organizations. Since OpenAI released ChatGPT in November 2022, blue chip companies have cautiously engaged the technology - seeking use cases which can deliver outcomes at scale. Such use cases have been identified in areas such as marketing and customer engagement, but less so in R&D (beyond personal productivity tools and novelty demonstrations). R&D leadership must learn (firsthand) what makes GenAI a distinct technology and then decide how best to harness its potential. Today, company Boards are looking past the GenAI hype and demanding value creation. In R&D, stories abound of promising experiments and applications, but few players in traditional industries have created significant value. Right now, companies are seeking competitive advantage through experimentation. What must come next are positive impacts on innovation outcomes and the ‘bottom line’.

Whether one is a GenAI convert or GenAI skeptic, all business leaders need to understand the technology and its potential impact on people and working practices. Whilst GenAI is part of a digital continuum, having much in common with earlier technologies, it is also materially different. Its disruptive potential is greater – changing not only the way we work, but also what it might mean to be an R&D professional in the years ahead. Although it may be tempting to take a ‘wait and see’ approach, GenAI is likely to fast become table stakes in R&D; delay risks lost advantage.

Many of us have started our journey along the GenAI maturity curve (a representation of which follows). However, without established roadmaps, most efforts rely on learning by doing. This Playbook aims to provide guidance for R&D leaders as they address the challenges and opportunities of GenAI. It shares the emerging insights and experience of companies in Sagentia’s CTO Forum.

In the sections that follows you will learn about the key stages of GenAI maturity observed by our Forum members. Within the three ‘Plays’, we offer ten principles that can help cross the chasm from early to established and successful use of GenAI. The adoption of GenAI will present many challenges for R&D leaders. We hope this Playbook helps you navigate the path successfully.

Maturity curve for GenAI adoption

Ten principles

Play 1: Fit with digital journey

1. Embrace the leadership challenge – consider GenAI as one element in your digital toolkit – one which places distinctive, new demands on R&D leadership

2. Commit to data – make smart data curation a central feature of your GenAI journey – leverage proprietary data to gain competitive advantage

3. Recognize the people challenge – expect people to move more slowly than the technology; understand the motivations and concerns of R&D professionals

4. Focus on value; be ready to iterate – prioritize value creation (rather than novelty) – be ready to reset expectations as your experience grows

Play 2: R&D Problems to solve

5. Find best fit – prioritize problems that are well suited to the idiosyncratic capabilities of GenAI and where resolution can deliver value, ideally for multiple team members

6. Manage for maturity – match your ambition to your organization’s GenAI maturity – do not run before you can walk 7. Distinguish between divergent and convergent tasks – understand how GenAI addresses distinct types of tasks – for both, keep humans in the loop

Play 3: Successful implementation of GenAI

8. Lead the change – develop programs to provide your people with motivation and training; ensure tools are simple to use

9. Activate specialist teams – form teams to provide leadership, guardrails and IT infrastructure; encourage pathfinders to identify and address R&D ‘problems to solve’

10. Protect skills, challenge role boundaries and respect specialisms – identify risks and opportunities linked to the introduction of GenAI; develop plans to manage them appropriately

Play 1: Fit with digital journey

Play 1 discusses how GenAI relates to organizations’ wider digital strategies. Whilst recognizing the similarities between GenAI and other digital tools, it calls out significant differences and factors that pose the greatest challenge for R&D leaders. It counsels R&D leaders to step in to the challenge, whilst retaining a focus on delivering business value. The overarching message from Play 1 is:

Strengthen foundations of data and people to drive value creation through GenAI.

Play 1 obeservations

Same but different: GenAI is a tool like any other (e.g. analytic AI) in R&D’s digital armory. Used well, it can help organizations deliver business goals. Success in GenAI requires familiar foundations – including data assets, technology, leadership, governance, and a workforce enabled and encouraged to use the technology. But beyond these similarities lie material differences that demand more from R&D leaders:

• Highly idiosyncratic technology using probabilistic not deterministic models

• A vast range of new use cases

• Reliance on many more data sources (both external and internal)

• Deployment by users across the R&D organization

• Very rapid rate of change (challenging for organizations and individuals)

• Unpredictable implications for R&D workflows and individual roles

In short, GenAI is poised to drive radical change in R&D. The pace of developments and the impact on R&D individuals are likely to be significant sources of disruption.

GenAI - a superpower for R&D: GenAI can elevate the ability of R&D employees. It connects people with multiple sources of data (internal and external) to create new pathways towards R&D goals. Through the power of Large Language Models, teams can access huge pools of proprietary data (which may have previously been untapped) to solve problems and create opportunities ‘faster, better, and cheaper’ than in the past. Proprietary data includes internal data and potentially data from customers, partners and suppliers.

Not there yet: Although GenAI promises much, use cases in R&D have generally delivered only incremental productivity benefits. Few are generating positive returns on investment; fewer still are delivering transformational change. Nevertheless, expectations remain high that GenAI will become table stakes for R&D organizations, and to realize the opportunity effectively it is important to embrace rather than defer adoption.

You can download a PDF of the entire playbook from Sagentia Innovation here, for free: https://sagentia.com/cto-forum/