Wildcard: Bring your nonprofit’s idea to life

If you already have an idea, start by defining what it needs to accomplish. If you need inspiration, think about the questions, reports, or repetitive tasks that regularly take up your team’s time.

You might build:

• A donor briefing with giving history and recent activity
• A campaign summary formatted for a board update
• A weekly list of lapsed donors who need follow-up
• A morning briefing highlighting recent gifts and campaign activity
• A simple fundraising dashboard

Choose something your team would realistically use, then define:

• What: What are you building?
• Who: Who will use it?
• Purpose: What should it help them understand or do?
• Format: What should the finished result look like?

Choose a compatible AI tool your team already uses, such as ChatGPT, Claude, Cursor, or Lovable. For Claude Code, Cursor, and other supported assistants, check the AI connector setup guidance for the latest options and instructions.

Inside your chosen AI tool, add the Givebutter connection. You’ll be directed to Givebutter to sign in and authorize access.

Review the requested access before approving it. Your existing Givebutter permissions determine which account information is available through the connection.

Once connected, your AI tool can work with the campaign, transaction, and supporter data available to you in Givebutter.

Before building your idea, confirm that the AI tool can find and accurately interpret your Givebutter data.

You can do this by asking a few questions you already know the answers to, such as:

• Which campaigns are currently active?
• How much has [campaign name] raised?
• Show me donations from the last 30 days.
• Who are our five largest recent donors?

Compare the responses with the information in Givebutter. If a result is incomplete or inaccurate, make the question more specific by naming the campaign, timeframe, transaction type, or other relevant criteria.

This quick check helps you understand which data is available and how precisely you need to ask for it. Remember that the Givebutter connection provides access to data; it does not allow the AI tool to make changes inside your Givebutter account.

Describe your idea as if you were assigning it to a teammate. Explain what you want to build, who will use it, what Givebutter data it should include, and how the result should be organized.

For example:

Build a simple weekly dashboard for our executive director. Show our active campaigns and, for the last seven days, the total amount raised, new donors, new recurring gifts, and five largest donations. Use our Givebutter data and make the dashboard easy to scan in under one minute.

Your first version could be a dashboard, donor briefing, weekly summary, research tool, or repeatable process. It does not need to include every possible feature.

Check important names, totals, dates, and conclusions against Givebutter before moving on. Save this first version so you can test and improve it in the next step.

Use the first version yourself, then ask someone else who would benefit from it to try it.

Find out:

• Does it answer the intended question?
• Is anything important missing?
• Is the result easy to understand?
• Would they use it again?

Use their feedback to improve your build. You might simplify the layout, add or remove a data point, narrow the timeframe, or change how the information is organized.

Check the revised result against Givebutter again. Once it is accurate and useful, save the final prompt and any instructions your team will need to run it again.