**Yes, there is a time dimension, and it has clear spikes, a long decline and a few data breaks.** The dataset covers roughly January 2015 to January 2020. The date column is stored as text, which causes some of the breaks below. **Deal count: plateau, then decline** - Deal volume rose from about 51–62 a month in early 2015 to a peak of 104 in January 2016. It was 86–98 a month from June 2015 to mid-2016. - It then drifted down through 2016, to 72 a month by late 2016. - It dropped to 48 in February 2017 and 43 in July 2017. It fell below 40 a month in 2018, to 25 in August. - The sharpest drop comes in autumn 2018: 13 deals in September, 7 in October, 18 in November and 7 in December. January 2019 has only 3. - Recorded activity in 2019 is thin, at 3–17 deals a month, with 17 in August 2019. Part of this may be incomplete coverage rather than a real market collapse, since deal counts here are only what was reported. **Funding amount spikes (total USD)** - **July 2015 is the biggest outlier.** One query put the month at about $9.8B, against typical months of $200M–$1B. A second query that parsed amounts differently gave about $1.53B for the same month, because it counted 62 amounts instead of 72. Treat the $9.8B as a few very large deals, or amount parsing artefacts, rather than a real month-wide surge. - Other spikes are 2015-09 (~$1.47B), 2017-03 (~$2.26B), 2017-05 (~$1.79B), 2017-08 (~$2.75B) and 2018-02 (~$1.21B). - In 2019 the totals spike in August (~$4.53B) and November (~$2.06B), even though deal counts there are only 17 and 15. These look like a few mega-deals. - The lowest recent months by total are 2018-10 (~$52.9M) and 2018-12 (~$57.7M), matching the deal-count collapse. **Median deal size is moving the other way** - Median ticket size rose while deal counts fell. It was about $0.4–1.0M in 2016, about $2–3.8M in late 2017 and $4–5M in several 2018 months. - It peaked at $30M in September 2018, on only 11 deals with amounts. It was $22M in January 2019. - So the market shifted toward fewer, larger deals. The small-deal tail of 2015–2016 shrank, although the late-2018 medians rest on very few deals. **Data breaks to be aware of** - **Malformed dates.** One deal is parsed as the year "0015" (July 2015, $0.63M). Seven more deals fall into a "-00" bucket (~$10.2M) with an unparseable month. - **Missing month.** March 2019 is absent from the results, which jump from 2019-02 to 2019-04. It may be a real gap or a date problem. - **Missing or text amounts.** About a third of deals have no usable amount. Deals with an amount run from roughly 55–75% of all deals in 2015–2016 to 80–100% in 2018. Some of the drop in totals in a given month may simply reflect undisclosed amounts. - **Sensitivity to parsing.** The two methods disagree for July 2015, and one gave NaN for July 2016. Amount-based spikes are therefore less reliable than deal counts. **Takeaway.** The reliable signals are the deal count peaking in early 2016 and falling steadily to a trough in late 2018, and the rising median ticket size. The individual spikes in total dollars come from a handful of very large deals and from amount-parsing artefacts.