**Overall, the number of reported deals grew through 2015, peaked in early 2016, then shrank steadily. Dollar totals are volatile and flat-to-spiky, with no stable seasonal pattern.** **Deal count (the clearest signal)** - Yearly deal counts fell every year after 2016: about 929–935 in 2015, 993 in 2016, 687 in 2017, 309–310 in 2018, 111 in 2019 and 7 in 2020. The 2020 figure covers only 2–13 January. - The monthly series shows the same shape. Deals rose from 51 in Jan 2015 to 97–98 in Jul–Aug 2015, peaked at 104 in Jan 2016, and slid to the 40s–50s by late 2017. In 2018 monthly deals fell from 37 in January to 7 in October and December, and 2019 had only 3 per month in January and February. - The 2018–2019 collapse is probably partly a data-collection effect, with fewer deals being reported or captured, rather than a pure market contraction. The dataset cannot separate the two. **Ticket size (rising)** - The median disclosed deal size rose from about $1.0M in 2016 to $1.5M in 2015's level, $2.25M in 2017, $3.9–4.0M in 2018 and $12M in 2019. - This rise is likely partly survivorship or reporting bias: as total deals thin out, the ones that get reported skew larger. The number of deals with a disclosed amount also fell (652 → 586 → 456 → 264 → 105). **Total dollars (spiky, no clear trend)** - Monthly totals are dominated by a few megadeals. Examples are Jul 2015 (~$1.53B), Mar 2017 (~$2.26B) and Aug 2017 (~$2.75B), against typical months of $200–500M. - Yearly totals therefore swing: about $10.4B in 2017, $5.1B in 2018 and $9.7B in 2019. A single megadeal can move any of these. - The 2015 and 2016 totals are unreliable. They differ between my queries (2015 ≈ $8.6B vs $16.9B, depending on how amounts were parsed), and 2016 returned `nan` in one version because of an unparseable amount in Jul 2016. I would not draw conclusions from them. **Seasonality** - There is no consistent seasonal cycle. The one visible seasonal-looking step is the jump from about 60 deals a month (Jan–May 2015) to about 90 from Jun 2015 onward. That is a level shift that persisted through 2015 and into 2016 rather than a repeating annual pattern. - Later years are dominated by the downtrend, so any seasonality is too weak to see. **Data quality** - The date column is stored as text. A few rows have malformed dates, such as a year recorded as "15" and some blank dates, which I left out of the yearly trend. The chart shows monthly deal counts, where the rise, the early-2016 peak and the decline are visible. {"action":"final","content":"","charts":[{"kind":"line","x":"month","series":["deals","deals_with_amount"],"title":"Monthly funding deals (2015–2019)"}]}