Thinking partners that transform AI weakness into sector strength

When the National Lottery Heritage Fund and Arts Marketing Association commissioned Goose in 2024, the initial brief seemed straightforward: create an AI assistant to help heritage organisations with their marketing challenges. However, extensive workshops with sector professionals revealed that what appeared to be a simple chatbot requirement was actually masking a far more complex strategic challenge.

The heritage sector's needs couldn't be met by an AI that provided definitive answers. Instead, they required something that had never been built before: an AI system that could help isolated professionals think more strategically by providing diverse perspectives rather than prescriptive solutions.

When "perfect answers" become the wrong solution

Early stakeholder conversations revealed a fundamental mismatch between typical AI capabilities and heritage sector needs. Workshop participants consistently described challenges that couldn't be resolved through information retrieval: "But it's just me," "Lack of support from leaders and trustees," "Smaller core team expected to deliver more with less."

These weren't knowledge gaps that could be filled by better access to marketing best practices. They were strategic thinking challenges faced by professionals who often worked in isolation, lacked diverse perspectives within their teams, and needed to navigate complex organisational dynamics while preserving heritage values.

The typical AI agency response would have been building a knowledge base system - templates, case studies, and frameworks that heritage professionals could consume. This would have completely missed the point. But our workshops revealed that generic solutions rarely translated effectively across the sector. A strategy that worked for the British Museum would need complete reimagining for a small maritime museum in Cornwall.

More fundamentally, we discovered that heritage professionals didn't need more answers - they needed better questions. The challenge wasn't accessing information but knowing what to ask in the first place.

Flipping the script on AI limitations

This insight led to a strategic pivot that transformed AI's supposed limitation into its greatest strength. Large language models are unreliable for providing definitive answers - they hallucinate, lack current information, and can't account for specific organisational contexts. But they excel at something far more valuable: role-playing different professional perspectives to help users think through problems more comprehensively.

We redesigned Goose around "thinking partners" - AI personas representing different expertise areas who could engage in strategic conversations rather than providing prescriptive advice. Instead of asking "What should our marketing strategy be?" users could explore "What questions should we ask about our audience development?" with personas representing market research specialists, board relations experts, or community engagement managers.

This approach preserved professional agency whilst addressing the isolation that many heritage professionals experience - refusing to replace human judgment with automated recommendations. A marketing manager working alone at a small heritage site could suddenly access perspectives equivalent to having a diverse team of specialists available for brainstorming and strategic guidance.

The thinking partners concept proved particularly powerful because it acknowledged the complexity and context-dependency of heritage marketing challenges. Rather than offering one-size-fits-all solutions, the AI helped professionals develop situational awareness and strategic thinking capabilities that could be applied across different contexts.

Building for professional empowerment, not replacement

The thinking partners approach reflected deeper values about AI's role in mission-driven organisations. We explicitly rejected the standard AI positioning of 'expert system that knows better than you' and instead built collaborative intelligence that enhanced rather than replaced professional judgment.

Each thinking partner was designed to ask probing questions rather than provide definitive recommendations. A market research agent might respond to "We need to attract more young people" by asking: "How are you defining 'young people'? What does your current data tell you about this group's engagement? What barriers might prevent them from visiting? What competing activities attract this demographic in your area?"

This questioning approach served multiple purposes. It helped professionals develop more sophisticated strategic thinking, ensured solutions remained contextually appropriate, and built organisational capability rather than dependency. Most importantly, it preserved professional agency - the AI enhanced human decision-making rather than replacing it.

For the National Lottery Heritage Fund, this represented a more sustainable approach to sector development. Rather than funding solutions that organisations would consume and forget, Goose was building transferable thinking skills that would strengthen the sector's long-term strategic capabilities.

From individual tools to collaborative ecosystems

The strategic pivot also revealed opportunities for peer learning that wouldn't have emerged from a traditional chatbot approach. While specific marketing tactics rarely transferred between organisations, strategic questions and thinking frameworks proved highly portable.

The alpha testing phase validated this approach. Usage data showed that projects incorporating thinking partners averaged 20 messages compared to 7 messages for standard conversations, suggesting users found genuine collaborative value rather than just information retrieval. User feedback consistently praised the sector-specific understanding: "Goose feels like I've asked a person and someone very knowledgeable has come back to me."

Perhaps most significantly, the beta phase launched with 40 heritage organisations across the UK, from major institutions to small community sites. The thinking partners approach scaled effectively across this diversity because good questions transfer even when specific answers don't.

The project timeline reflected this collaborative development approach. Rather than rushing to market with an impressive but potentially misaligned product, the team invested time in understanding sector needs, iterating on the strategic approach, and ensuring genuine utility. The January 2026 launch date allows for comprehensive beta testing and refinement based on real-world usage.

Strategic innovation for mission-critical contexts

Goose demonstrates that the most sophisticated AI applications for mission-driven organisations often emerge not from showcasing technological capabilities but from deep understanding of sector challenges and user needs. By transforming AI's limitation (inability to provide perfect answers) into its strength (ability to facilitate better thinking), the project created something genuinely innovative rather than simply applying existing AI patterns to heritage contexts.

The thinking partners concept represents a fundamentally different relationship between professionals and AI - one based on collaboration and empowerment rather than information consumption. For sectors facing resource constraints and professional isolation, this approach offers a pathway to enhanced capability without compromising the human expertise that defines mission-driven work.

The project's success will ultimately be measured not by user engagement or technical sophistication, but by whether heritage organisations develop stronger strategic thinking capabilities and feel more confident navigating complex challenges. This aligns perfectly with the National Lottery Heritage Fund's investment in sector sustainability and professional development rather than just technological solutions.