**The most surprising finding is that a handful of deals drive almost all the dollars, and the largest one looks like a data-entry error.** - **A few deals dominate.** Of 3,044 rows, 2,066 have a usable amount. In the clean aggregate (step 1, 2,065 valued deals), total funding is about $38.06B, but the median deal is only about $1.7M. The mean works out to roughly $18M, about 10 times the median. - **Deals of $100M or more carry most of the money.** The 73 deals of $100M+ are about 3.5% of valued deals but account for about 63.6% of all dollars. The other 96%+ of deals share the remaining third. - **The largest "deal" looks wrong.** The top row is Rapido Bike Taxi, Series B, on 27/08/2019, at $3.9 billion. That is about 10.2% of all funding in the dataset from one row. It is larger than Flipkart's $2.5B private equity round (11/08/2017) and Paytm's $1.4B round (18/05/2017). A Series B for a bike-taxi startup is out of line with those rounds. I haven't verified it against an outside source, but it looks like a units or decimal error, perhaps 1000× too large. - **Missing amounts add to the uncertainty.** 978 rows (about 32%) have no parseable amount. One of them, Drums Food on 21/07/2016, sorted to the top of my list with a null amount. **What this means:** - Any "total funding" or "average ticket size" built from this column is fragile. One suspect value moves the total by about 10%, and a few dozen mega-deals set the rest. - The median, or figures with the top deals excluded, is a more honest benchmark for a typical raise. - Several of my later aggregate queries returned NaN for totals and maxima, so I rely on the step-1 query for the dollar figures. The 2,065 vs 2,066 valued-row count differs by one between queries, probably the unparseable row, which is immaterial to the conclusion.