Use precise locations: When beginning a new chat, specify a zip code or city instead of asking about an unspecified area. For example, "What is the median income in zip code 64108?" is more precise than "What is the median income?"

An exception to this rule is when you have already told Sidekick your target location and want to ask a follow-up question. When this happens, Sidekick should be able to remember the provided location context in subsequent questions.

Be as specific as possible: If you need Sidekick to return a value, then be as specific as possible about the data you're interested in. Sidekick might struggle with broad topics — like poverty or the built environment — because there is so much data that it could draw from in our data library. Narrowing the topic can yield more satisfactory results. For instance, if you're asking about a broad topic like education, specify whether you're interested in graduation rates or school enrollment figures. This will increase the likelihood that Sidekick returns the information you want.

Ask a series of discrete questions rather than try to have Sidekick answer several questions at once: You're more likely to get the information that you're looking for if you can ask Sidekick individual, specific questions like "What is the diabetes rate in Jackson County, MO?" or "What is an age breakdown for 64131?" and then follow-up with more specific questions. Asking Sidekick a number of questions in a single query tends to yield less than satisfactory results.

Avoid ambiguity: Make sure your question is clear and direct. Ambiguous questions can lead to broad or unrelated answers. If you're not getting the response that you would expect from Sidekick, try reframing the question to be more specific.

Use complete sentences: While keywords like "health outcomes" or "housing costs" are important, framing them in a well-structured question helps the AI understand the context better.

Know the universe of possibilities: Understand that the AI pulls from Wherehouse, mySidewalk's data library. (Access the data catalog here.) Queries outside this library's scope might not return accurate results.

Current data only: At this time, Sidekick can only answer questions about the most current data available within our data library. However, we are working on making historical and projected data available.

Try framing your questions to return data recommendations: If you're not sure exactly what data applies to an issue you have in mind, try asking Sidekick to help you. For example, "Can you recommend data that might show how food access affects diabetes?"

Try variations: If you don't get the expected answer, rephrase your question or use different keywords.

Ask follow-up questions: If your initial query brings back general data, you can ask more detailed follow-up questions.

Your input matters: If you encounter issues or have suggestions, please provide feedback. This helps us improve Sidekick.