**Nine of the 13 columns have missing values: all nine business columns. The four `_ingestion_*`/`_batch_id`/`_source_*` metadata columns were not checked.** Stage is the worst, with about 28.5% blank across all 446 rows. The missing values are blank strings, not true NULLs. A NULL-only check (step 0) found nothing, and a NULL-only percentage (step 2) showed 0.0% everywhere. Counting blank or empty strings (steps 4–6) is what surfaces the gaps. Percentages below are the "ALL MONTHS" counts in step 6 divided by 446 rows. | Column | Missing rows | % missing | |---|---|---| | Stage | 127 | ~28.5% | | Investor | 45 | ~10.1% | | Founders | 38 | ~8.5% | | Amount | 25 | ~5.6% | | Location | 20 | ~4.5% | | Company Name | 19 | ~4.3% | | Sector | 19 | ~4.3% | | Founded | 18 | ~4.0% | | What it does | 18 | ~4.0% | The monthly breakdown (steps 4–6) shows two separate causes: - **Stage is blank mostly in January, February and March.** It is missing in 43 of 115 January rows (37.4%), 29 of 96 February rows (30.2%) and 31 of 98 March rows (31.6%). April is only 3.2% missing. May is about 50% missing. - **May is the main source of gaps in the other columns.** Company Name, Founded, Location, Sector and What it does are complete in January through April, apart from one blank Location in April. In May, 42–45% of rows (18–19 of 42) are blank in each. The dataset notes already flag that some May rows are misaligned or blank, so those are likely empty or broken rows rather than genuinely unknown values. - **Investor, Founders and Amount have smaller gaps in every month.** For example, April has 10 blank Founders (10.5%) and 10 blank Investors (10.5%), and May has about 50–64% blank in these. The January–April stage gaps look like genuinely undisclosed rounds, while the May gaps are mostly a data-quality problem. If you exclude May, most columns are essentially complete. Stage, Investor and Founders are the exceptions.