**The strongest relationship is between company age (`Founded`) and funding `Stage`.** Younger companies are at earlier stages. Stage is also the main driver of `Amount`. The correlations come from a pooled run across the five monthly tables, using the 426 rows (step 2). I turned Stage into an ordinal rank and used the natural log of Amount, because amounts are heavily skewed. | Pair | Pearson r | Rows used | |---|---|---| | `Founded` vs `Stage` | **-0.54** | 256 with a usable stage | | `Stage` vs ln(`Amount`) | **+0.40** | 385 with an amount, 256 with a stage | | `Founded` vs ln(`Amount`) | **-0.25** | 385 with an amount | - **`Founded` and `Stage`:** the average founding year rises from about 2003 for "undisclosed" stage and about 2012–2014 for Series C–F, to 2018–2019 for Series A and Seed, and 2021 for Pre-seed (step 1). - **`Stage` and `Amount`:** the median amount rises almost monotonically with stage. It is about $0.5M for Pre-seed, $1.6M for Seed, $12M for Series A, $60M for Series C, and $316.5M–$364M for Series F and E (step 1). The correlation is only moderate because the Series E median ($364M) sits slightly above Series F ($316.5M), and because of the wide spread within each stage. - **`Founded` and `Amount`:** older companies tend to raise larger rounds, but this is weaker. Most of it likely runs through stage, since older companies are at later stages. **Caveats** - **Amount is text.** It holds values like "$1,000,000" and "Undisclosed". "Undisclosed" is the most common value (35 rows), and 25 more rows are blank (step 0). The correlations only cover rows with a parseable amount. - **Stage labels are dirty.** There are about 30 distinct raw values, including variants such as `Preseed`, `Pre-seed`, `Undiclosed`, `$Series A` and `Series-D`. Some rows hold misaligned data, such as investor names in the Stage column, and some amounts are malformed (for example "$165.5" and "$728.100,000") (step 3). That is mostly the May table, which the card flags as misaligned. Those rows lower the reliability of the stage-based numbers. - **Categorical columns not tested.** I did not measure `Sector` or `Location` against Stage or Amount, so I can't say how strongly they relate. - **No joins.** The tables share no key, so these are pooled estimates, not cross-table links. ```json {"charts":[{"kind":"bar","x":"stage","series":["median_amount"],"title":"Median funding amount by stage"}]} ```