Skip to main content

Signal Check Views

Page views of each Signal Check in its first 14 days, and against a typical Signal Check at the same age.

How many times each Signal Check was viewed in its first 14 days, and whether a new one is ahead of or behind a typical Signal Check at the same age.

Always name a Signal Check as “Signal Check #N”, never “edition” or a bare “#N”.

Calculation

From meta_page_traffic, take the Signal Check pages: page_kind = 'signal_check' and a pagepath_redacted ending in <YYYYMMDD>_signal-check.html. The date in the file name is the day the Signal Check was sent.

  • Day 1 is the publication date. Views before it are the team reviewing the draft, so leave them out.
  • The first 14 days are the publication date and the 13 days after it. A fixed window makes Signal Checks comparable; views since publication favour the older ones.
  • Only Signal Checks published from 2026-07-23, when tracking began. Signal Check #1 and #2 came out before it and have no comparable figures.
  • The number comes from the page title, #N.

Typical Signal Check. To judge a new Signal Check that is, say, 5 days old, add up each earlier Signal Check’s views over its own first 5 days, and take the median. Use only Signal Checks with all 14 days tracked, and leave out the new one itself. Comparing at the same age is the point: a 5-day-old Signal Check against the others’ full 14 days always looks behind.

WITH pages AS (
    SELECT
        pagepath_redacted,
        CAST(date AS DATE) AS day,
        -- the date in the file name is the day it was sent
        CAST(date_parse(regexp_extract(pagepath_redacted, '([0-9]{8})_signal-check', 1), '%Y%m%d') AS DATE)
            AS published,
        CAST(regexp_extract(pagetitle, '#([0-9]+)', 1) AS INTEGER) AS number,
        screenpageviews
    FROM google_analytics.meta_page_traffic
    WHERE page_kind = 'signal_check'
      AND regexp_like(pagepath_redacted, '[0-9]{8}_signal-check[.]html$')
)

SELECT
    MAX(number) AS signal_check,
    published,
    -- day 1 is the publication date; earlier views are the draft being reviewed
    SUM(screenpageviews) FILTER (WHERE day BETWEEN published AND date_add('day', 13, published))
        AS views_first_14_days,
    SUM(screenpageviews) FILTER (WHERE day >= published) AS views_since_published
FROM pages
-- tracking began on 2026-07-23
WHERE published >= DATE '2026-07-23'
GROUP BY pagepath_redacted, published
ORDER BY published;
tbl(con, in_schema("google_analytics", "meta_page_traffic")) |>
  filter(
    page_kind == "signal_check",
    str_detect(pagepath_redacted, "[0-9]{8}_signal-check[.]html$")
  ) |>
  mutate(
    day = as.Date(date),
    # the date in the file name is the day it was sent
    published = sql("CAST(date_parse(regexp_extract(pagepath_redacted, '([0-9]{8})_signal-check', 1), '%Y%m%d') AS DATE)"),
    number = sql("CAST(regexp_extract(pagetitle, '#([0-9]+)', 1) AS INTEGER)")
  ) |>
  # tracking began on 2026-07-23
  filter(published >= as.Date("2026-07-23")) |>
  group_by(pagepath_redacted, published) |>
  summarise(
    signal_check = max(number, na.rm = TRUE),
    # day 1 is the publication date; earlier views are the draft being reviewed
    views_first_14_days = sum(
      if_else(day >= published & day <= sql("date_add('day', 13, published)"), screenpageviews, 0L),
      na.rm = TRUE
    ),
    views_since_published = sum(if_else(day >= published, screenpageviews, 0L), na.rm = TRUE),
    .groups = "drop"
  ) |>
  select(signal_check, published, views_first_14_days, views_since_published) |>
  arrange(published)

Interpretation notes

Views, not readers. A reader who opens a Signal Check twice counts twice. For how far readers get, see Dwell Time.

A young Signal Check is incomplete. Until its 14th day has been exported, views_first_14_days is still growing. Compare it with a typical Signal Check at the same age, not with the others’ 14-day totals.

Views come in early. Most of a Signal Check’s views land in the first few days, after it is announced, so a slow first day matters more than a slow tenth.

Alternative names

  • Signal Check readership
  • Signal Check reach
  • views per Signal Check
  • Signal Check page views