Take one real question from raw data all the way to a shareable answer, using everything from this module. You will submit the sharp question you wrote, a link to your working Google Sheet with the real dataset imported and cleaned, the one chart that shows the answer, your one-sentence insight in the 'X, which means Y, so we should Z' form, and a two-line note on the sanity check you ran and what could still be off. The test of success: a stranger reads your submission and understands the finding without you explaining a single thing.
Reuse the coffee-shop worked example as your template - your submission should have the same shape, just your data and question. When you paste data samples into AI to get formulas or column summaries, use only 10 sample rows and strip out any names or emails first. Set your Google Sheet share setting to 'Anyone with the link can view' so the reviewer can open it. If your chart is hard to read in five seconds, you probably have too many bars - cut it down to just the categories your question compares.
Question: Did customers from Instagram spend more per order than customers from Google in June? | Sheet: [link to Google Sheet, 240 rows, header frozen] | Chart: column chart titled 'Average order value by source, June' showing Instagram at 412 dollars and Google at 268. | Insight: 'June Instagram customers averaged 412 dollars per order versus 268 from Google - about 54 percent higher (X), which suggests Instagram is attracting higher-intent buyers rather than just more clicks (Y), so we should keep the Instagram campaign and test raising its budget 20 percent (Z).' | Sanity check: 'Summed all June revenue unfiltered and it matched the 31,000 dollars in our dashboard. Still uncertain: Instagram had only 38 orders vs Google's 202, so the average could swing with a few big orders.'