TechInsights modernises its content architecture for building smarter information products

Overview

TechInsights, a leading B2B Information Platform providing authoritative technical and market analysis for the semiconductor industry, faced significant challenges in content delivery. These included issues with speed, searchability, and discoverability of their valuable content. Their legacy systems, which relied on manual processes and static PDF documents, were hindering timely information access and limiting user engagement. By implementing Hygraph, TechInsights transformed (and not rebuild) its content architecture, resulting in faster delivery, an enhanced user experience, and a more data-driven approach to content creation.

The Challenge: Engaging Users with Content

Before adopting Hygraph, TechInsights' content infrastructure, while functional, struggled to keep pace with the fast-moving semiconductor industry. Customers required faster access to critical analysis for informed decision-making. Hussein El Aggan, TechInsights' Chief Architect, noted, "The feedback from our existing customers was that they love our content... but the problem is that they're not getting it fast enough in order to support their decision-making." As a platform with users, the primary KPI for TechInsights is to be laser-focused on Monthly Active Users (MAUs). Many of the bottlenecks had to do with their existing content operations, as outlined below:

Searchability and Discoverability 

Content was often difficult to locate, even when it existed. Users would frequently request analysis that TechInsights had already published. "Nine times out of ten, it was there. The problem was that users weren't finding the content that they were looking for," Hussein explained. "Then finding content once and then disappearing for six months does not help us achieve our internal targets." This lack of discoverability also impacted user engagement. 

Manual, Static, and Duplicative Processes 

Content creation involved manual steps, such as copying data into Excel and pasting screenshots into PDFs. This resulted in static content that required manual updates across multiple versions if a detail changed. "If we found something as innocuous as a typo in the executive summary, we had four or five places that we have to go and change it," Hussein highlighted the inefficiency.

Inconsistent Metadata and System Heterogeneity:

TechInsights used various disparate systems with differing or non-existent metadata capabilities. This inconsistency made centralized search and content relationships challenging without extensive custom development. Hussein provided an example: "All of a sudden, in order to just do something as simple as ‘I want to add a flag on a piece of content that says that it's free’, I had to have our software dev team implement six different methods."

Author Productivity and Focus 

Highly technical authors were spending valuable time on administrative tasks instead of focusing on their core analysis. "Our authors are deeply technical... They don't like writing executive summaries for our marketing website," Hussein stated.

The Solution: More modular and smarter 

TechInsights implemented Hygraph to create a flexible and interconnected content platform (See Figure Below). 

Modular Content Architecture: 

Content elements were structured as "lego blocks" within Hygraph's schema, allowing authors to assemble these blocks into reports while preserving data relationships. Hussein described the process as "What they're doing is they're taking all these lego blocks and they're putting them together to tell a story in the report."

Connected Data for Enhanced Discoverability: 

By retaining relationships between content elements, TechInsights built an "interconnected graph database." This enabled dynamic content recommendations and created entry points for users to navigate related information, where TechInsights describe it as "connected data”. It became a way for TechInsights to interconnect all of their content together. It also effectively allowed TechInsights avoiding a "rebuild" of its content architecture and instead federate content to fit the overall architecture (see image below).

Incremental Publishing: 

The new content operation allowed authors to publish content incrementally, releasing sections of a report as they were completed. This significantly improved perceived speed for customers and encouraged user return visits. Hussein observed, "Customers like it because they get it, gives them the perception of speed and it gives them incremental value."

Consistent Metadata and Automated Tagging: 

TechInsights defined a consistent set of minimum metadata for all content types and automated much of the tagging process. TechInsights is also exploring AI for auto-tagging where they write a simple mechanism that allows them to auto tag the content, essentially preventing the mess of inconsistent manual tagging. This allowed surfacing pages that are organised and more fitting to the experience that users consume content in their platform as shown below:

Dynamic Content Embedding: 

The platform now dynamically embeds content, using references instead of copies of images or dashboards. This ensures consistency and reduces manual updates.

The Results: Faster & Smarter

The implementation of Hygraph brought about significant improvements for TechInsights:

• Incremental updates vs. 6 weeks to publish (for 600-page reports): Modular content and incremental publishing allowed for quicker delivery of valuable information to customers. 

• Increased Monthly Active Users: The "perception of speed" and improved content discoverability encouraged users to return to the platform more frequently.

• Elimination of Duplicated Effort: Authors now produce a single version of content, with metadata controlling its visibility. Hussein emphasized, "If I update the executive summary for example, it automatically updates everywhere."

• Enhanced Productivity for Authors: Automation of summaries and tagging freed up technical experts to focus on core analysis.

• Improved User Experience: Dynamic content, interactive tables of contents, and related content suggestions made the platform more engaging and easier to navigate.

With the implementation taking one month to go live, an incremental approach, starting small and expanding gradually, helped TechInsights succeed where previous transformation attempts had failed. Within 6 months after go live, 80% of content was successfully migrated to Hygraph. Hussein confirmed, "We approached it in a very incremental way, and therefore we were able to convince people."

What’s next for TechInsights? AI, Data, and Personalisation 

TechInsights is planning to leverage AI (OpenAI and Bedrock) to generate content summaries and executive summaries, freeing authors from these tasks where future plans include persona-based content presentation.TechInsights sees content as data, so Hygraph can also provide valuable telemetry data on user engagement, informing future content strategy. Hussein explained, "We can start to feed that back into our content teams to say, look on these reports, you're getting a lot of engagement in these areas."

TechInsights' journey with Hygraph demonstrates how a strategic shift to a headless CMS can unlock new levels of efficiency, enhance user value, and transform an organization's approach to building information products even in the era of AI.

Interested to see how TechInsights shape the future of content directly in action? Jump to techinsights.com and follow their latest updates and innovation.