Hey everyone! I'm new here and diving deep into the analytics stack at my company. We've been trialing Granola for a few weeks, and I'm really excited about its promise, especially the "automated forecasting" feature they highlight.
I've been trying to evaluate it beyond the marketing gloss. In my tests, it seems to automatically detect seasonality in our sales data, which is pretty cool. But I'm left wondering about the "black box" feeling. When I compare its forecasts to some manual ARIMA models I've built, the results are... different. Not necessarily worse, but I don't fully understand *why*.
Could anyone with more experience walk me through:
* What's actually happening under the hood? Is it just a packaged Prophet model, or something more?
* How do you validate its forecasts? What's a good process to build trust in it?
* Are there specific data shapes or scenarios where it really shines (or falls flat)?
* For a beginner in forecasting, is this a good tool to learn concepts, or does it hide too much?
I'm also curious about integrating these forecasts into our Looker dashboards. Has anyone set up a pipeline where Granola's forecasts are automatically fed into a BI tool for comparison against actuals? 😊
Any practical advice or "gotchas" would be super helpful!