**Overall trend: deal activity grew into 2016, then shrank steadily. Dollar totals are volatile, with no clear direction. Seasonality is weak.** **Deal counts (reliable)** - Yearly deal counts (step 3/4): 2015 had about 929–935 deals and 2016 peaked at 993. After that: 687 in 2017, 309–310 in 2018, and 111 in 2019. 2020 has only 7 deals, all in early January, so it is a partial year. - Monthly counts show the same pattern. Counts climbed from 51 in Jan 2015 to a peak of 104 in Jan 2016. They then slid to the 40s by late 2017 and to single digits or low teens by late 2018 (for example 7 in Oct 2018 and 7 in Dec 2018). - The 2019 count of 111 deals does not match the monthly rows I could see. Jan and Feb 2019 had only 3 deals each, so the rest of 2019 must be much denser. The monthly query was capped at 50 of 60 rows, so I did not see those months. Together with the jump in median ticket size, this suggests the late records come from a different collection or reporting regime, so the decline may partly reflect how deals were captured rather than real activity. **Deal size** - The median disclosed ticket rises over time: about $1.5M in 2015, $1.0M in 2016, $2.25M in 2017, about $3.9–4.0M in 2018, $12M in 2019 and $9M in the 2020 rows. Fewer deals are being reported, but the ones that are reported tend to be larger. **Total dollars (noisy, treat with caution)** - Totals jump around month to month, driven by a few mega-deals. Examples are about $2.26B in Mar 2017, $2.75B in Aug 2017, $1.53B in Jul 2015, and $1.47B in Sep 2015. Neighbouring months are often a few hundred million. - Annual totals are inconsistent across my queries. 2015 comes out at about $8.6B in step 4 but about $16.9B in step 3, which suggests different amount-parsing rules. The 2016 total came back as NaN because at least one amount could not be parsed. The totals therefore do not support a firm trend, though 2017 (about $10.4B) and 2019 (about $9.7B) look higher than 2016 and 2018. **Seasonality** - There is no consistent seasonal pattern. In 2015, deal counts stepped up from about 60 per month in Jan–May to about 90–98 from Jun–Dec. That step did not repeat in 2016 or 2017, so it looks like part of the growth phase rather than a seasonal cycle. **Caveats** - One row has a malformed date and falls in a "year 15" bucket, and 7 rows have no date at all. - The amount column is missing for about a third of deals in 2015–2017, so dollar figures cover only the deals that disclosed an amount.