Research Leadership Network - Generative AI set to transform outsourced business research with 20-50% efficiencies
In recent conversations with clients in global leadership positions in business research, we’ve discussed the impact of generative AI the Knowledge Process Outsourcing (KPO) business model.
None of the leaders we’ve spoken to are inputting client data into open systems or relying on raw outputs from generative AI for client deliverables. Generative AI is more often being applied to internal knowledge bases or vendor data acting as a copilot. Primarily the biggest impacts seem to be when generative AI is used for speeding up the aggregation and presentation of information from multiple sources. It is also useful when challenging researchers to think about ‘what else’ when framing a research question.
- 20-50% efficiencies reported from several KPO sources, particularly on tasks relating to speeding up writing and formatting answers for basic research requests.
- Huge potential to increase creativity and productivity with the reduction in time required to experiment with research questions.
Challenges:
- Research leaders are focusing on stakeholder management and influencing up within their organisations. Leadership teams at the executive level are getting pitched daily by emerging vendors with AI solutions making bold claims.
- Big risk that unobvious contextual inaccuracies are missed through hallucinations, this is a key opportunity and focus for experienced researchers.
- Are outsourced KPO’s resilient enough to face any copyright breach incidents?
- Are KPO’s training their own models with client data and queries?
In organisations with large research functions there are layers to a research workflow from external data providers to external KPO’s, internal research teams and fee earners.
Which layers are best placed to create value with these new tools and what is the model that creates the most value moving forward?