Personalisation

Q: What have you learnt about using AI in personalisation?

A: AI-driven personalisation offers publishers the opportunity to engage and add value for their audiences, by tailoring their experience with a publisher’s content to their individual needs. AI allows publishers to provide individual users with targeted information that aligns with their specific interests, roles, or industries, thereby improving the overall user experience and engagement levels.

Our experience developing MyCatalyst, an AI-powered chat discovery tool tailored for the analyst research industry, has provided several key insights into how AI can best serve publishers in personalising user experiences. One of the key learnings is that a familiar, intuitive user interface is essential. With the rapid adoption of AI tools like ChatGPT, users are becoming accustomed to certain design conventions and functionality. Keeping the chat interface familiar by adopting widely understood structures while tailoring them to the brand and content tone enables users to navigate the platform confidently and intuitively.

Another key takeaway is the importance of respecting content licenses within personalised recommendations. For instance, if a user searches for insights on European coffee trends but only has a license for tea-related data, an AI tool should acknowledge the relevant coffee report’s existence without granting access. This approach supports transparency while encouraging potential upgrades, ensuring AI-driven suggestions align with licensing and commercial considerations.

We’ve also learnt to deal with a few of the pitfalls of AI, for example protecting against incorrect answers. While it’s impossible to protect entirely against hallucinations, guard rails, like beta testing and a phased roll-out, must be in place to mitigate risk.

Q: In which use-case has AI proved most effective?

A: A particularly effective use-case for AI-driven personalisation has been in the implementation of AI chat discovery. This allows publishers to deliver customised and interactive content recommendations in real time, providing an engaging, conversational interface that automatically surfaces relevant content based on user queries. This allows users to find insights more intuitively, as the chat function not only responds to questions but also proactively suggests relevant resources.

Compared to conventional search, AI search is more intelligent and will give you one detailed answer to your question, drawn from multiple sources. Importantly, it can also show you exactly where to find the references it used, whether on a web page, in a Word doc, presentation, or PDF, at a page level.

For instance, a user looking for recent developments in their sector could type a query into the chat. Based on this input, the AI can summarise the publisher’s content in this area and recommend specific reports for further insights. This allows users to access relevant insights without the need for extensive searching, saving them time and making large databases of content more accessible. It also offers up-sell opportunities by recommending relevant content that may be available outside a user’s existing subscription.

The ROI of AI is likely to be measured in terms of productivity, customer and employee satisfaction, and retention / renewals but will depend on an organisation’s business objectives. We’ve calculated that AI-powered search offers significant added value for publishers from a time and cost savings perspective.

Three best practice top tips

1. Use an intuitive AI-powered chat interface to serve up personalised content recommendations from your content library. Users today are familiar with popular AI chat tools, so personalisation interfaces should feel intuitive and easy to navigate. By adopting recognisable layouts and user-friendly navigation elements, publishers can help users focus on finding insights rather than learning a new tool. Additionally, an AI chat function makes your whole library discoverable and offers opportunities for upsell.
2. Ensuring AI recommendations respect user licenses is crucial. An effective AI chat or content personalisation tool can identify what users have access to and provide results accordingly. This approach improves user satisfaction while highlighting additional content that may interest them.
3. Make AI-generated responses transparent and include legal disclaimers. This transparency helps users understand the nature of the insights they receive and prepares them for the potential limitations of automated responses. Additionally, incorporating a visible legal disclaimer within an AI chat interface that users must acknowledge can further protect both publishers and end users.

Content Catalyst is the first-choice technology partner for ambitious analyst research firms. Using our AI-powered technology, publishers of analyst reports and data can launch subscriber portals that maximise the value of their insights and drive revenues.

Website: www.contentcatalyst.com

Tel: 01223 617 322

LinkedIn: www.linkedin.com/in/ben-tregenna-72b774