This deck sets out how AI trend analysis would be used inside IRG’s stakeholder impact study. The study itself asks how organizations successfully pivot from shareholder primacy to stakeholder impact, and the AI work is one of the methods feeding that question.
Key findings
- The study has five focus areas: defining stakeholder impact, identifying the initiatives that make the most impact, measuring and communicating impact, the growth leadership profile, and the implications for each key role.
- The analysis runs on two bodies of text: interview transcripts, and public posts on LinkedIn and Twitter scraped against a keyword list.
- The keyword set is deliberately small. Six substantive words are tracked, together with the four C-suite roles the study addresses.
- The central comparison is between overperformers and underperformers, and specifically what the two groups talk about differently.
- The output is framed as a prompt rather than a report. The intended deliverable tells a leader what they are not talking about and should be.
Three phases across a year
The roadmap runs in three phases. The first, from January to May, is interviews and research, made up of vision interviews and a meta-analysis of published initiatives.
The second phase, from May to September, produces preliminary results: topline conclusions, a preliminary report, and a sparring session with the advisory board.
The third phase, from October to December, is the full publication. It carries best practices and case studies, frameworks and practical tools, and success measures.
What the graph work is looking for
The first pass ranks the six keywords for overperformers and underperformers and maps what sits around each word. The plan calls this the neighborhood: not the term itself, but what it pulls in with it.
A second study restricts the same graph by the speaker’s position in the top leadership team, which surfaces the differences across the functions rather than across performance.
The contrast between the two groups is then handled with a hypergraph approach that works across things, connections, and concepts. The instruction attached to the method is to use the data as it exists.
Where the AI work sits among the other methods
The AI analysis is not the whole study. It sits alongside peer group events and a partner sparring session, which are listed as parallel contents of the program.
Scope and limitations
This is a plan rather than a result. The deck describes methods, phases and intended outputs, and contains no findings from the analysis it proposes.
Source
This page summarises AI Trendanalysis, by Institute for Real Growth, Saïd Business School, Oxford Future of Marketing Initiative, May.