使用数据工作室中大查询的数据重新创建谷歌分析仪表板

Question

I am currently trying to recreate my google analytics dashboard using the big query connector within ga4. I am using a custom query to pull the data i need from Big Query and display it in data studio. When i just calculate some KPIs from the data and pull those using the query, the information comes through correct. But once I try to access custom event data, i have to unnest it before it is accessible. Once unnested, the unnested data needs to be grouped or the query will throw an error. The grouping of rows seems to be messing up the KPI's i previously calculated and inflates the values. How do i query this data correctly? I tried to pull all the raw data in and use custom fields within data studio to calculate the fields i need but ran into issues there too.

SELECT
distinct
  event_date,
  event_timestamp,
  event_name,
  user_pseudo_id,
  device.category,
  -- author
  (
  SELECT
  distinct
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='author'
    group by 1 ) AS author,
  -- campaign
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='campaign' 
    group by 1 ) AS campaign,
  -- categories
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='categories' 
    group by 1 ) AS categories,
  -- clientid
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='clientid'
    group by 1  ) AS clientid,
  -- duration
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='duration'
    group by 1  ) AS duration,
  -- eventactions
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='eventactions'
    group by 1  ) AS eventactions,
  -- eventcategory
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='eventcategory'
    group by 1  ) AS eventcategory,
  -- eventlabel
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='eventlabel'
    group by 1  ) AS eventlabel,
  -- mediatype
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='mediatype'
    group by 1  ) AS mediatype,
  -- pagetitle
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='pagetitle' 
    group by 1 ) AS pagetitle,
  -- pagetype
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='pagetype' 
    group by 1 ) AS pagetype,
  -- source
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='source'
    group by 1  ) AS SOURCE,
  -- sourceurl
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='sourceurl' 
    group by 1 ) AS sourceurl,
  -- srclink
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='srclink' 
    group by 1 ) AS srclink,
  -- status
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='status'
    group by 1  ) AS status,
  -- title
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='title' 
    group by 1 ) AS title,
  -- user_clientid
  (
  SELECT
    params.value.string_value
  FROM
    UNNEST(event_params) AS params
  WHERE
    params.key='user_clientid' 
    group by 1 ) AS user_clientid,
  traffic_source.source AS User_Source,
  -- end groupby
  COUNT(1) AS eventCount,
  SAFE_DIVIDE(COUNT(DISTINCT
      CASE
        WHEN ( SELECT value.string_value FROM UNNEST(event_params) WHERE key = 'session_engaged') = '1' THEN CONCAT(user_pseudo_id,( SELECT value.int_value FROM UNNEST(event_params) WHERE key = 'ga_session_id'))
    END
      ),COUNT(DISTINCT CONCAT(user_pseudo_id,(
        SELECT
          value.int_value
        FROM
          UNNEST(event_params)
        WHERE
          key = 'ga_session_id')))) AS engagement_rate,
  COUNT(DISTINCT user_pseudo_id) AS Unique_Users,
  COUNT(DISTINCT
    CASE
      WHEN ( SELECT value.int_value FROM UNNEST(event_params) WHERE key = 'engagement_time_msec') > 0 OR ( SELECT value.string_value FROM UNNEST(event_params) WHERE key = 'session_engaged') = '1' THEN user_pseudo_id
    ELSE
    NULL
  END
    ) AS active_users,
  COUNT(DISTINCT
    CASE
      WHEN ( SELECT value.int_value FROM UNNEST(event_params) WHERE key = 'ga_session_number') = 1 THEN user_pseudo_id
    ELSE
    NULL
  END
    ) AS new_users,
  COUNT(DISTINCT user_pseudo_id) AS users
FROM
  `zngly-corporate.analytics_315869392.events_*`,
  UNNEST(event_params) AS params
GROUP BY
  1,
  2,
  3,
  4,
  5,
  6,
  7,
  8,
  9,
  10,
  11,
  12,
  13,
  14,
  15,
  16,
  17,
  18,
  19,
  20,
  21,
  22,
  23

Above is the sql query that I have tried. Not sure how i can achieve what i need correctly, if anyone has any insight to share I would appreciate it.

Answer 1

The SQL for this is quite complex. You might try a query builder like Analytics Canvas to verify your result set: https://analyticscanvas.com/ga4-bigquery-query-builder/ . There's a free trial, so you don't need to buy the software to verify your results and get you back on track.

Be super careful when connecting Looker Studio to BQ directly. It sends separate queries for each chart, table, scorecard and user of the report. If you connect directly, ensure you are creating date partitioned summary tables and hitting those, instead of the raw GA4 export tables.