Thought leadership disrupted:
a manifesto for change

AI may be devaluing content production, but it is also creating new, innovative research methods that can displace the traditional survey-based model. Data that fuels thought leadership campaigns can now come from more or less anywhere: earnings calls, analyst reports, social media sentiment, job postings and many other sources. Consider a company that wants to look at M&A intentions in a sector for the year ahead. The traditional approach would be to survey corporate development and finance executives to ask them directly. But today, there are other much more efficient ways of getting this information in a way that is robust, accurate and reliable. And crucially, these new methods are much quicker than the traditional eight-week survey cycle.

Other AI-powered methods are also emerging, albeit outside the mainstream. Some are using AI-powered voice agents to conduct primary research by calling thousands of venues to establish a price index, for example. Elsewhere, researchers are testing how multiple AI agents can be configured to forecast economic trends. In these frameworks, various agents analyse real-time data—text, charts, and news—to predict market shifts. While these methods are still evolving, the broader takeaway is clear: thought leadership professionals must remain open to new, AI-driven methodologies.

While many insights professionals are exploring AI-generated ‘synthetic data’ to reduce survey costs, questions remain regarding the quality and validity of such samples. Instead, one emerging option is the use of AI-based conversational platforms. These platforms use AI agents to conduct qualitative interviews at scale, synthesising hundreds of conversations in days. While currently focused on consumers rather than niche B2B executives, the method is faster, cheaper, and richer than traditional approaches. Respondents are encouraged to explain their reasoning and provide nuance through non-linear dialogue. Even where structured surveys remain necessary, providers are using AI to accelerate every stage, from design and targeting to final analysis.

The first improvements come from how surveys are designed and how samples are chosen. Most surveys are limited in scope by budget. Rather than draw conclusions from limited samples, producers should be transparent about methodology shortcomings, and admit that certain findings may still need further research before they can be considered as irrefutable conclusions.

Sampling bias can also be a problem, particularly if companies are surveying their own clients. Look at the way that survey questionnaires are positioned. Are they more likely to attract respondents who have strong views about a subject, or particular demographic groups? If so, look at the language and make it more neutral to appeal to a wider sample.

We explained in the first section that confusion between correlation and causation is a common problem in thought leadership surveys. It is just too tempting to conclude that performing a certain action - investing in a certain technology, diversity and inclusion or artificial intelligence, for example - leads to strong performance. If two findings are correlated, by all means say so, but again be open about the possibility that this could be due to a variety of different factors, including chance. Correlations can be interesting, but they remain hypotheses rather than conclusions.

The term narrative fallacy, coined by the author Nassim Taleb, refers to the tendency to create simple stories to explain complex events. This is very common in thought leadership reports, because producers often want to provide an explanation for certain results. That is commendable, because interpretation is useful to the audience, but it can be problematic if that interpretation makes connections that aren’t really there or ignores the effects of other important factors. So explore different interpretations of findings and be open that alternative explanations may be possible. Simple narratives are appealing but they’re not that helpful if they are wrong.

Shifting to web-native, modular layout platforms can eliminate a lot of this friction. Writers can drop copy into pre-designed templates that adjust automatically and dynamically as text is revised. Development also no longer needs to be a bottleneck. AI front-end generators (such as v0, Lovable, or Claude Artifacts) allow teams to paste a content brief or wireframe prompt and create production-ready code in minutes, entirely bypassing traditional web development backlogs. Finally, standing out from the crowd requires breaking the traditional approaches and adopting more dynamic, engaging, and modular formats. 

Surprisingly, relatively few B2B thought leadership campaigns feature interactive or dynamic elements. This represents an untapped opportunity. Interactive tools and data visualisations, such as self-assessments, quizzes, benchmarks, data dashboards or similar, all help to turn passive readers into more active participants. And rather than serving up static conclusions, more interactive formats can also allow executives to query the underlying data through the lens of their own company, sector, or geography. 

Providing a range of content formats also helps; everyone has different preferences for how they consume content. Video and/or audio content and smart data visualisation all help to replace dense blocks of text with more engaging insights. 

The most successful modern campaigns do not treat derivative assets (such as infographics, podcasts, or social carousels) as an afterthought. Instead, they design a multi-format ecosystem from day one. Shifting from static assets to more dynamic, interactive, and modular experiences, B2B brands can better differentiate themselves, while also improving their impact on AI search. 

The typical approach to securing this input is to book in a 60-minute call. But what if this approach could be rethought, with the executive able to provide short voice notes or WhatsApp messages in spare minutes that can be collected, aggregated and analysed in a much more efficient way? This removes the bottleneck while still accessing the necessary insight.

Rather than seeing AI as a bolt-on cosmetic tool to enhance existing processes, a more fundamental approach means rethinking workflows from the ground up, building multi-agent systems that can process steps concurrently rather than relying on a lengthy linear process. Consider, for example, the content development stages. This typically relies on a core asset that is developed first (usually a long-form report) with various spin-off content assets such as videos, articles and presentation decks being produced afterwards. Often, a last-minute change to copy on one asset means reworking the others, too - a tedious, lengthy, manual sequence. Instead of this linear process, an asset builder agent enables parallel production. It acts as an orchestrator, loading LLMs with raw manuscripts and brand design guidelines to automatically generate a variety of web native publications. Last-minute changes can be updated on the core manuscript, and flow through automatically to every content asset.

This means moving beyond marketing metrics, such as downloads, MQLs and dwell time to something that CFOs will value: pipeline and revenues. “There’s often too much of an MQL mindset with an emphasis on gated content and leads,” says Joel Harrison. “There needs to be tangible, quantifiable results from any marketing activity, including thought leadership.”

Contribution to revenues can be challenging, however, in sectors where sales cycles may extend to years rather than months. Proxies are sometimes needed alongside hard revenue metrics, which means that a balance is needed to prevent a slide towards short-termism and an over emphasis on converting, rather than creating, demand. “We need proxies that might not be hard dollars, but that directionally show that new value has been created,” says Brian Macreadie, Head of Marketing at Addleshaw Goddard. “This reflects a long-term focus and an understanding that things don’t sell overnight. “Meaningful conversations might not convert immediately but people understand that they influence future pipeline.”