A Conversation with Dan Ridsdale
1) How has AI changed the way you structure or present your research outputs from data visualization to copy and to report formatting?
With ChatGPT now driving more traffic than LinkedIn, AI is redefining how we shape and share our research. Since only a small proportion of LLM users click through to source links, our research is reaching far more readers than direct traffic implies.
Our core value proposition is to build bigger, better-informed investor audiences for our clients, and we recognised early on that discoverability is just as important as distribution, so we’re embracing AI in the same way as we did SEO. Increasingly, content is being “atomised” and put together alongside data from other AI sources. As a result, companies need to prioritise the quality and structure of their data and narrative, and we definitely believe we are a key piece of that jigsaw.
Most importantly, our research remains open access, so it can be referenced by search engines and AI’s. Other basics, such as good meta-tagging remain important for AI search as much as search engines. For certain types of content, we are also structuring question-based headings (similar to FAQs) to ensure that we’re being referenced for the most frequently asked questions over AI. Ultimately, AI prioritises content that delivers the most value to users, so our focus on producing insightful, high-quality research remains unchanged. Credibility also matters—having an analyst’s name and the strength of our brand enhances authority signals.
AI is also expanding how we repurpose our analysis. We’re leveraging tools to transform research into multiple formats—summaries, visuals, and multimedia content—tailored to different investor audiences. For example, when we initiate coverage on a company, we also create thematic pieces that provide additional discovery pathways for investors focused on that sector or trend.
2. What steps do you take to validate or cross-check AI-generated insights to ensure analytical accuracy and reliability?
Our research is always written and reviewed by humans, both the supervisory analyst and editorial levels and that will not change. We are, however, equipping analysts with capable and compliant AI tools to enhance and accelerate the research writing process. On the review side, we’re applying greater scrutiny to ensure every output can be traced back to a verifiable source, while also checking for bias and maintaining an engaging human tone. At the same time, we are increasingly using AI to make the review process more efficient, highlighting inconsistencies in numbers, forecasts etc. while verifying sources, and also identifying areas where the text is not easy to read.
3. Looking ahead, which aspects of the research process do you think hold the most untapped potential for AI integration and why?
For us, the equation is quite different than for the buy-side or even many broker firms. While we have many high volume and quant funds ingesting our research, our fundamental proposition is about providing good quality research that enables human investors to develop a strong understanding of our clients and their investment case. One of the biggest wins is simply reducing the drudgery tasks such as plugging in numbers manually, wordsmithing etc. to free up our analysts for more value added work. AI tools can also dramatically accelerate the research process when used correctly. We’re also using AI to build bespoke, targeted investor lists for our clients and to enhance engagement analytics, giving them deeper insight into who is engaging with their equity story.
