Why your charity website needs to be a museum, not just a library

We recently hosted a webinar with our partners at Platypus Digital to discuss how charities can adapt to the rise of AI search.

The arrival of tools like ChatGPT, Gemini and Google’s AI Overviews has caused a lot of anxiety in the sector. People are worried about algorithms, disappearing traffic and how to keep their organisations relevant.

But the answer isn’t to chase the algorithm. The answer is to double down on human-centred design.

To survive this shift, charities need to change how they view their websites. They need to stop acting like libraries and start acting like museums.

Reconsidering the library model

For a long time, charity websites have functioned like libraries. They have been repositories of facts, figures and quick answers.

If a user wanted to know the symptoms of a condition or how to donate, they searched for it, clicked a link, found the fact and left.

AI search is now taking over this function.

If you ask Google or ChatGPT for a simple fact, it will give you the answer directly on the results page or in the chat interface. The user gets what they need without ever clicking through to your website.

If your content strategy is based on providing quick, functional answers, AI will cut you out of the journey. As such, you need to revisit the balance of your content strategy.

Becoming a museum

If the library is for looking up facts, the museum is for immersive experiences.

Museums offer stories, deep context and emotional connection. You go to a museum not just to know something, but to feel something and explore.

Charities need to pivot their content to offer what AI cannot:

• Deep expertise: Insight that comes from years of experience, not just scraping the web.

• Human stories: Real voices from beneficiaries and frontline staff that build empathy.

• Trust: A verified, safe source of truth in an era of AI hallucinations.

This is the “Museum” model. It is about creating content that is so rich and distinct that a summary generated by a robot simply isn’t enough.

The exception for health charities

There is a major exception to this rule: medical information.

For health charities, reputation and accuracy are critical. Users—and search engines—still prioritise trusted sources for medical advice because AI tools are prone to making things up.

In this space, being the “library” is still valuable. However, you must ensure your content is structured in a way that AI can read and cite it, ensuring your organisation remains the source of truth.

Traffic will drop, and that is fine

This shift requires a change in mindset regarding metrics. We are proactively telling clients that their traffic will likely drop.

AI acts as a filter. It strips away the low-intent users who just wanted a definition or a quick date.

The traffic that remains is higher intent. These are people looking for help, connection or a way to get involved. They are more valuable to your cause than the thousands who previously bounced after ten seconds.

We need to move away from vanity metrics and focus on measuring genuine engagement. These are metrics like time on page, pages per visit, engagement rate etc - not to mention working harder to test our pages with our audiences.

Practical ways to adapt

You cannot beat AI at generating average content. To create “Museum” quality content, you need to leverage your human advantages.

Co-design with your audience

AI can summarise what has already been written, but it cannot empathise with what people are feeling right now. To create content that truly connects, you need to get out of the building.

Spend time with your service users. Look at what they are saying on social media and what they are asking your customer enquiry lines.

Don’t just guess what they need. Get people in a room, show them examples of your content and ask them what works. This direct human insight is your biggest competitive advantage against a generic algorithm.

Use paired writing

AI generates generic text. To stand out, you need to extract the deep knowledge that lives inside your organisation.

We advocate for “paired writing”. This brings a communications expert and a subject matter expert together. The comms person draws the story out, and the expert ensures the nuance and depth are there. The result is distinct content that an LLM could never produce on its own.

Build a community of practice to create rapid feedback loops

SEO strategies used to be static documents owned by a single team. That doesn’t work anymore.

AI is changing too fast for one person to track. You need to build a community of practice that cuts across your organisation—bringing together colleagues from fundraising, policy and services.

Use this group to create rapid feedback loops. Share what you are seeing in search results and what clients are telling you. If ChatGPT is giving outdated answers about your services, this group can spot it and act immediately.

This moves you from a “set and forget” strategy to a culture of shared learning and quick response.

The litmus test

If you are unsure where your content stands, try this simple test:

Ask ChatGPT a question that your charity answers.

If the AI gives a good, complete answer, your page is a “Library” page. It is at risk of dying. You need to evolve it by adding stories, video, unique data or tools.

If the AI cannot answer it fully, or if the answer lacks the human nuance you provide, you are on the right track.