As we continue to increase the quality and value of research provided to clients, particularly by integrating AI-assisted innovation, we’re introducing a new value-add: data visualizations. Creating these data visualizations typically should take approximately 10 -15 minutes (less as the process becomes more familiar!)
Data visualizations are graphic representations of data that help people understand patterns and trends, identify relationships between variables, and spot outliers. This is true even for complex datasets. Visual elements such as charts, graphs, and maps are used to present information in an accessible and understandable way.
<aside>
<img src="/icons/view_blue.svg" alt="/icons/view_blue.svg" width="40px" /> We will include our data visualization directly below the Summary section.
</aside>
Using ChatGPT for Data Visualizations
Directions may be found here.
<aside>
<img src="/icons/exclamation-mark-double_blue.svg" alt="/icons/exclamation-mark-double_blue.svg" width="40px" /> You may use this Data Visualization Wizard, and it will do most of the work for you! Start out by typing in the chat box:
"I need a visualization for this research: [copy-paste research content]" in one prompt.
When copy-pasting, exclude anything other than the research + research strategy (not proposed next steps or the template language re: sprints).
</aside>
Tools Needed
- Advanced-Data Analysis in your ChatGPT Plus account
- A free Canva account
- This Canva template
- Follow these steps to create each visualization if you would rather type in the prompt instead of using the GPT shared above.
Ways to Make ChatGPT Visualizations More Effective
- When analyzing data in a spreadsheet, the model may have difficulty determining the correct data visualization and accurately reflecting the data if the data is not presently clean within the sheet. To avoid this, keeping your spreadsheets as clean as possible is helpful. Meaning: keep key data points separate from explanations of that data (e.g. if providing revenues for 10 companies and this required you to convert currencies to U.S. dollars, put the $ value in one cell and explanation in another - this helps the AI model easily identify the key data.
- If the data is not presented cleanly, Advanced Data Analysis can clean it for you and produce a new spreadsheet to work from.
- Take a look at how pasting only relevant sections can affect the visualization.
- For extremely long research reports, ChatGPT may not be able to process the entire research if pasted within the prompt. If this is the case, select the most relevant to the client, given the context of the project. If that data does not yield a logical visualization, try that section of the research that is best suited for a visualization (i.e. focuses on quantitative data).
- Learn Basic Data Visualization Principles: Understanding the basics of data visualization will help you choose the right type of chart and make effective design decisions.
- Be Specific in Your Requests: Use precise terms and be clear about what changes you want to see.
- Provide Sample Visuals: If possible, share examples of visuals that align with your vision.
- Iterate and Provide Feedback: Visualization is often an iterative process. Provide feedback on drafts to refine the visual.