51黑料不打烊

文档Experience Platform仪表板指南

配置文件分析

Last update: Fri Apr 04 2025 00:00:00 GMT+0000 (Coordinated Universal Time)

创建对象：

  • 开发人员
  • 用户

通过分析数据模型而获得的见解，可使您的51黑料不打烊 Real-Time CDP数据更易于访问、理解并影响决策。

通过访问为配置文件提供支持的SQL来了解配置文件洞察，然后生成您自己的洞察以进一步探索您的客户及其构成您配置文件的消费者体验。 通过使用现有的Real-Time CDP数据模型SQL作为灵感，根据独特的业务需求创建查询，将原始数据转换为新的可操作洞察。

有关如何直接通过Experience Platform UI调整您的分析的SQL的更多信息，请参阅查看厂蚕尝文档。

以下分析均可用作用户档案仪表板或自定义用户定义仪表板的一部分。 有关如何自定义仪表板或?在构件库和用户定义的仪表板中创建和编辑新构件的说明，请参阅自定义概述。

按合并策略列出的受众重迭 audience-overlap-by-merge-policy

通过此洞察回答的问题：

  • 哪些配置文件对两个受众通用？
  • 重迭对参与率或转化率有何影响？
  • 如何为重迭的区段定制营销策略？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT Sum(overlap_col1) overlap_col1,
        Sum(overlap_col2) overlap_col2,
        Sum(overlap_count) Overlap_count
  FROM
    (SELECT 0 overlap_col1,
            0 overlap_col2,
            sum(count_of_overlap)Overlap_count
    FROM qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments
    WHERE qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments.merge_policy_id = 2027892989
      AND qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments.date_key = '2024-01-10'
      AND ((qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments.segment1=1333234510
            AND qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments.segment2=1559754729)
            OR (qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments.segment1=1559754729
                AND qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments.segment2=1333234510))
    UNION ALL SELECT sum(count_of_profiles) overlap_col1,
                      0 overlap_col2,
                      0 overlap_count
    FROM qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines
    LEFT JOIN qsaccel.profile_agg.adwh_dim_segments ON qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.segment_Id = qsaccel.profile_agg.adwh_dim_segments.segment_Id
    WHERE qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.merge_policy_id = 2027892989
      AND qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.date_key = '2024-01-10'
      AND qsaccel.profile_agg.adwh_dim_segments.segment_Id = 1333234510
    UNION ALL SELECT 0 overlap_col1,
                      sum(count_of_profiles) overlap_col2,
                      0 Overlap_count
    FROM qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines
    JOIN qsaccel.profile_agg.adwh_dim_segments ON qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.segment_Id = qsaccel.profile_agg.adwh_dim_segments.segment_Id
    WHERE qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.merge_policy_id = 2027892989
      AND qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.date_key = '2024-01-10'
      AND qsaccel.profile_agg.adwh_dim_segments.segment_Id = 1559754729 ) a;

有关此分析的外观和功能的信息，请参阅按合并策略构件划分的受众重迭文档。

受众重迭报告 audience-overlap-report

通过此洞察回答的问题：

  • 50个最重迭的受众是什么？
  • 50个最不重迭的受众是哪个？
  • 重迭模式如何因合并策略而变化？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT source_segment_name,
        source_segment_id,
        overlap_segment_name,
        overlap_segment_id,
        max(source_segment_audience_count) source_segment_audience_count,
        max(overlap_segment_audience_count) overlap_segment_audience_count,
        max(overlap_audience_count) overlap_audience_count,
        CASE
            WHEN (max(source_segment_audience_count) + max(overlap_segment_audience_count) - max(overlap_audience_count)) > 0 THEN (cast(max(overlap_audience_count) AS DECIMAL(18, 2)) / cast((max(source_segment_audience_count) + max(overlap_segment_audience_count) - max(overlap_audience_count)) AS DECIMAL(18, 2))) * 100::DECIMAL(9, 2)
            ELSE 100.00
        END overlapping_percentage
  FROM
    (SELECT adwh_fact_profile_overlap_of_segments.Segment1 source_segment_id,
            adwh_fact_profile_overlap_of_segments.Segment2 overlap_segment_id,
            Sum(count_of_overlap) overlap_audience_count
    FROM qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments
    WHERE qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments.merge_policy_id = 2027892989
      AND qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments.date_key = '2024-01-10'
    GROUP BY qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments.Segment2 ,
              qsaccel.profile_agg.adwh_fact_profile_overlap_of_segments.Segment1) a
  INNER JOIN
    (SELECT sum(count_of_profiles) source_segment_audience_count,
            adwh_dim_segments.segment_name source_segment_name,
            adwh_fact_profile_by_segment_trendlines.merge_policy_id,
            adwh_fact_profile_by_segment_trendlines.segment_Id segment1
    FROM qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines
    JOIN qsaccel.profile_agg.adwh_dim_segments ON qsaccel.profile_agg.adwh_dim_segments.segment_id = qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.segment_Id
    WHERE qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.merge_policy_id = 2027892989
      AND qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.date_key = '2024-01-10'
    GROUP BY qsaccel.profile_agg.adwh_dim_segments.segment_name,
              qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.merge_policy_id,
              qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.segment_id) b ON a.source_segment_id = b.segment1
  INNER JOIN
    (SELECT sum(count_of_profiles) overlap_segment_audience_count,
            adwh_dim_segments.segment_name overlap_segment_name,
            adwh_fact_profile_by_segment_trendlines.merge_policy_id,
            adwh_fact_profile_by_segment_trendlines.segment_Id segment2
    FROM qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines
    JOIN qsaccel.profile_agg.adwh_dim_segments ON adwh_dim_segments.segment_id = adwh_fact_profile_by_segment_trendlines.segment_Id
    WHERE qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.merge_policy_id = 2027892989
      AND qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.date_key = '2024-01-10'
    GROUP BY qsaccel.profile_agg.adwh_dim_segments.segment_name,
              qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.merge_policy_id,
              qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.segment_id) c ON a.overlap_segment_id = c.segment2
  GROUP BY source_segment_name,
          source_segment_id,
          overlap_segment_name,
          overlap_segment_id
  ORDER BY overlapping_percentage DESC
  LIMIT 5;

有关此分析的外观和功能的信息，请参阅受众重迭报表构件文档。

受众（计数） audiences

通过此洞察回答的问题：

  • 哪种合并策略主要用于分段？
  • 受众在合并策略中的分布情况如何？
  • 随着时间的推移，特定合并策略的受众数量是否有任何重大变化？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT count(DISTINCT a.segment_id) count_of_segments
  FROM qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines a
  JOIN
    (SELECT MAX(process_date) last_process_date,
            merge_policy_id
    FROM qsaccel.profile_agg.adwh_lkup_process_delta_log
    WHERE process_name = 'FACT_TABLES_PROCESSING'
      AND process_status = 'SUCCESSFUL'
    GROUP BY merge_policy_id) b ON a.merge_policy_id= b.merge_policy_id
  AND a.date_key = b.last_process_date
  WHERE a.merge_policy_id= 2027892989;

有关此分析的外观和功能的信息，请参阅受众小组件文档。

映射到目标状态的受众 audiences-mapped-to-destination-status

通过此洞察回答的问题：

  • 受众在映射目标和未映射目标之间的总体分布情况如何？
  • 哪些特定目标的映射受众数量最多？
  • 未映射的总受众比例如何？
  • 在这些未映射的受众中，是否存在模式或相关趋势？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT COUNT(DISTINCT (y.segment_id)) AS count_mapped_segments,
        COUNT(DISTINCT (x.segment_id)) - COUNT(DISTINCT (y.segment_id)) AS count_unmapped_segments,
        COUNT(DISTINCT (x.segment_id)) AS total_segments
  FROM qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines x
  LEFT JOIN qsaccel.profile_agg.adwh_dim_br_segment_destinations y ON x.segment_id = y.segment_id
  INNER JOIN
    (SELECT MAX(process_date) last_process_date,
            merge_policy_id
    FROM qsaccel.profile_agg.adwh_lkup_process_delta_log
    WHERE process_name = 'FACT_TABLES_PROCESSING'
      AND process_status = 'SUCCESSFUL'
    GROUP BY merge_policy_id) z ON x.merge_policy_id = z.merge_policy_id
  AND x.date_key = z.last_process_date
  WHERE x.merge_policy_id = 2027892989;

有关此分析的外观和功能的信息，请参阅映射到目标状态小组件文档。

受众规模 audiences-size

通过此洞察回答的问题：

  • 哪个受众区段规模最大？
  • 最大的五个受众是什么？
  • 主要受众的受众规模分布会随着时间发生何种变化？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.date_key,
        qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_name,
        qsaccel.profile_agg.adwh_dim_segments.segment,
        qsaccel.profile_agg.adwh_dim_segments.segment_name,
        sum(qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.count_of_profiles)count_of_profiles
  FROM qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines
  LEFT OUTER JOIN qsaccel.profile_agg.adwh_dim_segments ON qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.segment_id = qsaccel.profile_agg.adwh_dim_segments.segment_id
  LEFT OUTER JOIN qsaccel.profile_agg.adwh_dim_merge_policies ON qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.merge_policy_id=adwh_dim_merge_policies.merge_policy_id
  WHERE qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.date_key = '2024-01-10'
    AND qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.merge_policy_id= 2027892989
  GROUP BY qsaccel.profile_agg.adwh_fact_profile_by_segment_trendlines.date_key,
          qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_name,
          qsaccel.profile_agg.adwh_dim_segments.segment,
          qsaccel.profile_agg.adwh_dim_segments.segment_name
  ORDER BY count_of_profiles DESC
  LIMIT 20;

有关此分析的外观和功能的信息，请参阅受众大小小组件文档。

客户人工智能得分分布 customer-ai-distribution-of-scores

通过此洞察回答的问题：

  • 我的每个Customer AI模型在各存储桶中的得分分布情况如何？
  • 按高、中、低得分分布情况如何？
  • 按合并策略划分得分分布的划分是什么？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT b.model_name,
     b.model_type,
     CASE
         WHEN score >= 0
              AND score < 25 THEN 'LOW'
         WHEN score >= 25
              AND score < 75 THEN 'MEDIUM'
         WHEN score >= 75
              AND score <= 100 THEN 'HIGH'
     END bucket_name,
     CASE
         WHEN score >= 0
              AND score < 5 THEN '02.50'
         WHEN score >= 5
              AND score < 10 THEN '07.50'
         WHEN score >= 10
              AND score < 15 THEN '12.50'
         WHEN score >= 15
              AND score < 20 THEN '17.50'
         WHEN score >= 20
              AND score < 25 THEN '22.50'
         WHEN score >= 25
              AND score < 30 THEN '27.50'
         WHEN score >= 30
              AND score < 35 THEN '32.50'
         WHEN score >= 35
              AND score < 40 THEN '37.50'
         WHEN score >= 40
              AND score < 45 THEN '42.50'
         WHEN score >= 45
              AND score < 50 THEN '47.50'
         WHEN score >= 50
              AND score < 55 THEN '52.50'
         WHEN score >= 55
              AND score < 60 THEN '57.50'
         WHEN score >= 60
              AND score < 65 THEN '62.50'
         WHEN score >= 65
              AND score < 70 THEN '67.50'
         WHEN score >= 70
              AND score < 75 THEN '72.50'
         WHEN score >= 75
              AND score < 80 THEN '77.50'
         WHEN score >= 80
              AND score < 85 THEN '82.50'
         WHEN score >= 85
              AND score < 90 THEN '87.50'
         WHEN score >= 90
              AND score < 95 THEN '92.50'
         WHEN score >= 95
              AND score <= 100 THEN '97.50'
     END score_bins,
     Sum(CASE
             WHEN score >= 0
                  AND score < 25 THEN count_of_profiles
             WHEN score >= 25
                  AND score < 75 THEN count_of_profiles
             WHEN score >= 75
                  AND score <= 100 THEN count_of_profiles
         END) count_of_profiles
  FROM qsaccel.profile_agg.adwh_fact_profile_ai_models a
  JOIN qsaccel.profile_agg.adwh_dim_ai_models b ON a.merge_policy_id = b.merge_policy_id
  AND a.model_id = b.model_id
  WHERE a.merge_policy_id = 2027892989
    AND a.model_id = 1829081696
    AND score_date =
      (SELECT Max(score_date)
       FROM qsaccel.profile_agg.adwh_fact_profile_ai_models d
       WHERE d.model_id = a.model_id) GROUP  BY b.model_name,
          model_type,
          CASE
              WHEN score >= 0
                   AND score < 25 THEN 'LOW'
              WHEN score >= 25
                   AND score < 75 THEN 'MEDIUM'
              WHEN score >= 75
                   AND score <= 100 THEN 'HIGH'
          END,
          CASE
              WHEN score >= 0
                   AND score < 5 THEN '02.50'
              WHEN score >= 5
                   AND score < 10 THEN '07.50'
              WHEN score >= 10
                   AND score < 15 THEN '12.50'
              WHEN score >= 15
                   AND score < 20 THEN '17.50'
              WHEN score >= 20
                   AND score < 25 THEN '22.50'
              WHEN score >= 25
                   AND score < 30 THEN '27.50'
              WHEN score >= 30
                   AND score < 35 THEN '32.50'
              WHEN score >= 35
                   AND score < 40 THEN '37.50'
              WHEN score >= 40
                   AND score < 45 THEN '42.50'
              WHEN score >= 45
                   AND score < 50 THEN '47.50'
              WHEN score >= 50
                   AND score < 55 THEN '52.50'
              WHEN score >= 55
                   AND score < 60 THEN '57.50'
              WHEN score >= 60
                   AND score < 65 THEN '62.50'
              WHEN score >= 65
                   AND score < 70 THEN '67.50'
              WHEN score >= 70
                   AND score < 75 THEN '72.50'
              WHEN score >= 75
                   AND score < 80 THEN '77.50'
              WHEN score >= 80
                   AND score < 85 THEN '82.50'
              WHEN score >= 85
                   AND score < 90 THEN '87.50'
              WHEN score >= 90
                   AND score < 95 THEN '92.50'
              WHEN score >= 95
                   AND score <= 100 THEN '97.50'
          END;

有关此洞察的外观和功能的信息，请参阅得分构件的Customer AI分发文档。

客户人工智能评分汇总 customer-ai-scoring-summary

通过此洞察回答的问题：

  • 我的每个Customer AI模型的得分摘要是什么？
  • 我的客户人工智能倾向分数如何针对不同受众发生变化？
  • 与用户档案概述中的其他碍笔滨相比，我的得分摘要有何变化？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT model_name,
         model_type,
         CASE
             WHEN score BETWEEN 0 AND 24 THEN 'LOW'
             WHEN score BETWEEN 25 AND 74 THEN 'MEDIUM'
             WHEN score BETWEEN 75 AND 100 THEN 'HIGH'
         END score_buckets,
         sum(count_of_profiles) count_of_profiles
  FROM QSAccel.profile_agg.adwh_fact_profile_ai_models a
  JOIN QSAccel.profile_agg.adwh_dim_ai_models b ON a.merge_policy_id=b.merge_policy_id
  AND a.model_id=b.model_id
  WHERE a.merge_policy_id=2027892989
    AND a.model_id =1829081696
    AND score_date=
      (SELECT max(score_date)
       FROM QSAccel.profile_agg.adwh_fact_profile_ai_models d
       WHERE d.model_id=a.model_id)
  GROUP BY model_name,
           model_type,
           CASE
               WHEN score BETWEEN 0 AND 24 THEN 'LOW'
               WHEN score BETWEEN 25 AND 74 THEN 'MEDIUM'
               WHEN score BETWEEN 75 AND 100 THEN 'HIGH'
           END;

有关此洞察的外观和功能的信息，请参阅客户人工智能评分摘要构件文档。

身份标识重迭 identity-overlap

通过此洞察回答的问题：

  • 标识类型础和标识类型叠之间的公共交叉点是什么？
  • 如何根据特定身份类型之间的重迭来优化客户受众，以增强有针对性的营销策略？
  • 评估交叉区域内的营销活动绩效可以获得哪些见解？
  • 利用这一营销活动效果洞察，如何优化未来的营销工作？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT Sum(overlap_col1) overlap_col1,
        Sum(overlap_col2) overlap_col2,
        coalesce(Sum(overlap_count), 0) overlap_count
  FROM
    (SELECT 0 overlap_col1,
            0 overlap_col2,
            Sum(count_of_profiles) overlap_count
    FROM qsaccel.profile_agg.adwh_fact_profile_overlap_of_namespace
    WHERE qsaccel.profile_agg.adwh_fact_profile_overlap_of_namespace.merge_policy_id = 2027892989
      AND qsaccel.profile_agg.adwh_fact_profile_overlap_of_namespace.date_key = '2024-01-10'
      AND qsaccel.profile_agg.adwh_fact_profile_overlap_of_namespace.overlap_id IN
        (SELECT a.overlap_id
          FROM
            (SELECT qsaccel.profile_agg.adwh_dim_overlap_namespaces.overlap_id overlap_id,
                    count(*) cnt_num
            FROM qsaccel.profile_agg.adwh_dim_overlap_namespaces
            WHERE qsaccel.profile_agg.adwh_dim_overlap_namespaces.merge_policy_id = 2027892989
              AND qsaccel.profile_agg.adwh_dim_overlap_namespaces.overlap_namespaces in ('avid',
                                                                                          'crmid')
            GROUP BY qsaccel.profile_agg.adwh_dim_overlap_namespaces.overlap_id)a
          WHERE a.cnt_num>1 )
    UNION ALL SELECT count_of_profiles overlap_col1,
                      0 overlap_col2,
                      0 overlap_count
    FROM qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines
    JOIN qsaccel.profile_agg.adwh_dim_namespaces ON qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.namespace_id = qsaccel.profile_agg.adwh_dim_namespaces.namespace_id
    AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id = qsaccel.profile_agg.adwh_dim_namespaces.merge_policy_id
    WHERE qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id = 2027892989
      AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.date_key = '2024-01-10'
      AND qsaccel.profile_agg.adwh_dim_namespaces.namespace_description = 'avid'
    UNION ALL SELECT 0 overlap_col1,
                      count_of_profiles overlap_col2,
                      0 Overlap_count
    FROM qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines
    JOIN qsaccel.profile_agg.adwh_dim_namespaces ON qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.namespace_id = qsaccel.profile_agg.adwh_dim_namespaces.namespace_id
    AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id = qsaccel.profile_agg.adwh_dim_namespaces.merge_policy_id
    WHERE qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id = 2027892989
      AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.date_key = '2024-01-10'
      AND qsaccel.profile_agg.adwh_dim_namespaces.namespace_description = 'crmid' )a;

有关此分析的外观和功能的信息，请参阅身份重迭构件文档。

轮廓计数 profile-count

通过此洞察回答的问题：

  • 51黑料不打烊 Real-Time Customer Data Platform中的总配置文件计数是多少？
  • 如何根据合并策略分发用户档案？
  • 哪个合并策略具有最高的配置文件计数？

生成这些分析的厂蚕濒如下所示：

SELECT qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_name,
       sum(qsaccel.profile_agg.adwh_fact_profile.count_of_profiles) CNT
  FROM qsaccel.profile_agg.adwh_fact_profile
  LEFT OUTER JOIN qsaccel.profile_agg.adwh_dim_merge_policies ON qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_id=adwh_fact_profile.merge_policy_id
  WHERE qsaccel.profile_agg.adwh_fact_profile.date_key='2024-01-10'
    AND qsaccel.profile_agg.adwh_fact_profile.merge_policy_id = 2027892989
  GROUP BY qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_name;

有关此洞察外观和功能的完整信息可在配置文件计数构件指南中找到。

有关此分析的外观和功能的信息，请参阅配置文件计数构件文档。

轮廓计数变化 profile-count-change

通过此洞察回答的问题：

  • 总体配置文件计数变化的趋势是什么？
  • 哪些原因导致配置文件计数显着激增或骤减？
  • 是否有特定合并策略可驱动配置文件计数更改？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT (sum(count_of_profiles) - sum(count_of_profiles_days_ago)) profiles_added
  FROM
    (SELECT sum(qsaccel.profile_agg.adwh_fact_profile.count_of_profiles) count_of_profiles,
            0 count_of_profiles_days_ago
    FROM qsaccel.profile_agg.adwh_fact_profile
    WHERE qsaccel.profile_agg.adwh_fact_profile.merge_policy_id = 2027892989
      AND qsaccel.profile_agg.adwh_fact_profile.date_key = '2024-01-10'
    UNION ALL SELECT 0 count_of_profiles,
                      CASE
                          WHEN sum(cntondatediff) =0 THEN sum(cntmin)
                          ELSE sum(cntondatediff)
                      END AS count_of_profiles_days_ago
    FROM
      (SELECT coalesce(sum(qsaccel.profile_agg.adwh_fact_profile_by_trendlines.count_of_profiles), 0) cntondatediff,
              0 cntmin
        FROM qsaccel.profile_agg.adwh_fact_profile_by_trendlines
        WHERE qsaccel.profile_agg.adwh_fact_profile_by_trendlines.merge_policy_id =2027892989
          AND qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key =dateadd(DAY, - 30, '2024-01-10')
        UNION ALL SELECT 0 cntondatediff,
                        sum(qsaccel.profile_agg.adwh_fact_profile_by_trendlines.count_of_profiles) countMin
        FROM qsaccel.profile_agg.adwh_fact_profile_by_trendlines
        WHERE qsaccel.profile_agg.adwh_fact_profile_by_trendlines.merge_policy_id = 2027892989
          AND qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key =
            (SELECT min(qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key) col
            FROM qsaccel.profile_agg.adwh_fact_profile_by_trendlines
            WHERE qsaccel.profile_agg.adwh_fact_profile_by_trendlines.merge_policy_id =2027892989
              AND qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key >= dateadd(DAY, - 30, '2024-01-10')
              AND qsaccel.profile_agg.adwh_fact_profile_by_trendlines.count_of_profiles IS NOT NULL) )b) a;

有关此分析的外观和功能的信息，请参阅配置文件计数更改构件文档。

配置文件计数变化趋势 profile-count-change-trend

通过此洞察回答的问题：

  • 根据合并策略，过去12个月内用户档案计数变化的总体趋势是什么？
  • 过去30天内用户档案计数是否有特定的变化模式或波动需要引起注意？
  • 与总体趋势相比，过去90天内用户档案计数有何变化？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT date_key,
         profiles_count_change
  FROM
    (SELECT rn_num,
            date_key,
            (count_of_profiles-lag(count_of_profiles, 1, 0) over(
                                                            ORDER BY date_key))profiles_count_change
    FROM
      (SELECT qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key,
              sum(qsaccel.profile_agg.adwh_fact_profile_by_trendlines.count_of_profiles) count_of_profiles,
              row_number() OVER (
                              ORDER BY qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key) rn_num
      FROM qsaccel.profile_agg.adwh_fact_profile_by_trendlines
  WHERE qsaccel.profile_agg.adwh_fact_profile_by_trendlines.merge_policy_id = 2027892989
    AND qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key >=dateadd(DAY, - 30 -1, '2024-01-10')
  GROUP BY qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key)a)b
  WHERE rn_num > 1;

有关此分析的外观和功能的信息，请参阅配置文件计数更改趋势构件文档。

轮廓计数趋势 profile-count-trend

通过此洞察回答的问题：

  • 过去30天内基于合并策略的配置文件计数的整体趋势是什么？
  • 根据此趋势，与长期趋势（例如，90天和12个月）有何异同？
  • 在指定的时间段（30天、90天和12个月）内，哪个合并策略对配置文件计数增加或减少的贡献最大？
  • 用户档案计数中是否有任何特定的峰值或峰值与30天时间范围内的某些事件或期间相关联？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT date_key,
       sum(count_of_profiles) AS count_of_profiles
  FROM qsaccel.profile_agg.adwh_fact_profile_by_trendlines x
  INNER JOIN
    (SELECT MAX(process_date) last_process_date,
            merge_policy_id
     FROM qsaccel.profile_agg.adwh_lkup_process_delta_log
     WHERE process_name = 'FACT_TABLES_PROCESSING'
       AND process_status = 'SUCCESSFUL'
     GROUP BY merge_policy_id) y ON x.merge_policy_id = y.merge_policy_id
  WHERE date_key >= dateadd(DAY, -365, y.last_process_date)
    AND x.merge_policy_id = 2027892989
  GROUP BY date_key;

有关此分析的外观和功能的信息，请参阅配置文件计数趋势构件文档。

按身份标识列出的轮廓 profiles-by-identity

通过此洞察回答的问题：

  • 在配置文件总数中，哪种身份类型所占的比例更高？
  • 身份类型之间是否存在显着差异？
  • 身份类型的总体分布情况如何？
  • 身份计数是否存在任何显着差异或异常？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT qsaccel.profile_agg.adwh_dim_namespaces.namespace_description,
        sum(qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.count_of_profiles) count_of_profiles
  FROM qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines
  LEFT OUTER JOIN qsaccel.profile_agg.adwh_dim_namespaces ON qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.namespace_id = qsaccel.profile_agg.adwh_dim_namespaces.namespace_id
  AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id = qsaccel.profile_agg.adwh_dim_namespaces.merge_policy_id
  WHERE qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id = 2027892989
    AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.date_key = '2024-01-10'
  GROUP BY qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.date_key,
          qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id,
          qsaccel.profile_agg.adwh_dim_namespaces.namespace_description
  ORDER BY count_of_profiles DESC;

有关此分析的外观和功能的信息，请参阅按身份构件列出的配置文件文档。

轮廓计数变化趋势 profiles-count-change-trend

通过此洞察回答的问题：

  • 根据合并策略，过去12个月内用户档案计数变化的总体趋势是什么？
  • 过去30天内用户档案计数变化是否存在特定的模式或波动需要引起注意？
  • 与总体趋势相比，过去90天内用户档案数量的变化如何？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT date_key,
         profiles_count_change
  FROM
    (SELECT rn_num,
            date_key,
            (count_of_profiles-lag(count_of_profiles, 1, 0) over(
                                                            ORDER BY date_key))profiles_count_change
    FROM
      (SELECT qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key,
              sum(qsaccel.profile_agg.adwh_fact_profile_by_trendlines.count_of_profiles) count_of_profiles,
              row_number() OVER (
                              ORDER BY qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key) rn_num
      FROM qsaccel.profile_agg.adwh_fact_profile_by_trendlines
  WHERE qsaccel.profile_agg.adwh_fact_profile_by_trendlines.merge_policy_id = 2027892989
    AND qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key >=dateadd(DAY, - 30 -1, '2024-01-10')
  GROUP BY qsaccel.profile_agg.adwh_fact_profile_by_trendlines.date_key)a)b
  WHERE rn_num > 1;

有关此分析的外观和功能的信息，请参阅配置文件计数更改趋势构件文档。

按身份标识分类的轮廓计数变化趋势 profiles-count-change-trend-by-identity

通过此洞察回答的问题：

  • 过去12个月中不同身份配置文件计数变化的总体趋势是什么？
  • 是否有任何特定的身份趋势显示过去30天内发生了显着变化？
  • 比较特定身份的30天、90天和12个月趋势时，用户档案计数的变化有何不同？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT date_key,
        namespace_description,
        profiles_count_change
  FROM
    (SELECT rn_num,
            date_key,
            namespace_description,
            (count_of_profiles - lag(count_of_profiles, 1, 0) over(PARTITION BY namespace_description
                                                                  ORDER BY date_key)) profiles_count_change
    FROM
      (SELECT qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.date_key,
              qsaccel.profile_agg.adwh_dim_namespaces.namespace_description,
              sum(qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.count_of_profiles) count_of_profiles,
              row_number() OVER (PARTITION BY qsaccel.profile_agg.adwh_dim_namespaces.namespace_description
                                  ORDER BY qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.date_key) rn_num
        FROM qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines
        LEFT OUTER JOIN qsaccel.profile_agg.adwh_dim_namespaces ON qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.namespace_id = qsaccel.profile_agg.adwh_dim_namespaces.namespace_id
        AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id = qsaccel.profile_agg.adwh_dim_namespaces.merge_policy_id
        WHERE qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id = 2027892989
          AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.namespace_id= -1042977439
          AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.date_key >= dateadd(DAY, - 30 -1, '2024-01-10')
        GROUP BY qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.date_key,
                adwh_dim_namespaces.namespace_description)a)b
  WHERE rn_num > 1;

有关此分析的外观和功能的信息，请参阅按身份构件列出的配置文件计数变化趋势。

单一身份标识轮廓 single-identity-profiles

通过此洞察回答的问题：

  • 我的客户身份数据是否一致地用单个身份表示？
  • 我的用户群中有多少百分比只包含单一身份类型的配置文件？
  • 对于仅具有单一身份类型的配置文件，这会如何影响配置文件完整性？
  • 最常见的身份类型与单个身份配置文件计数之间是否存在关联？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_name,
       sum(qsaccel.profile_agg.adwh_fact_profile.count_of_Single_Identity_profiles) CNT
  FROM qsaccel.profile_agg.adwh_fact_profile
  LEFT OUTER JOIN qsaccel.profile_agg.adwh_dim_merge_policies ON qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_id=adwh_fact_profile.merge_policy_id
  WHERE qsaccel.profile_agg.adwh_fact_profile.date_key='2024-01-10'
    AND qsaccel.profile_agg.adwh_fact_profile.merge_policy_id = 2027892989
  GROUP BY qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_name;

有关此分析的外观和功能的信息，请参阅单一身份配置文件小部件文档。

按身份标识列出的单一身份标识轮廓 single-identity-profiles-by-identity

通过此洞察回答的问题：

  • 有多少独特客户使用单一身份（例如，电子邮件或电话号码）注册？
  • 单一身份配置文件在不同身份类型（如电子邮件或电话号码）之间的分布情况如何？
  • 单一身份配置文件中是否存在新兴的身份模式或转变？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT qsaccel.profile_agg.adwh_dim_namespaces.namespace_description,
        sum(qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.count_of_Single_Identity_profiles) count_of_Single_Identity_profiles
  FROM qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines
  LEFT OUTER JOIN qsaccel.profile_agg.adwh_dim_namespaces ON qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.namespace_id = qsaccel.profile_agg.adwh_dim_namespaces.namespace_id
  AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id = qsaccel.profile_agg.adwh_dim_namespaces.merge_policy_id
  WHERE qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id = 2027892989
    AND qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.date_key = '2024-01-10'
  GROUP BY qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.date_key,
          qsaccel.profile_agg.adwh_fact_profile_by_namespace_trendlines.merge_policy_id,
          qsaccel.profile_agg.adwh_dim_namespaces.namespace_description;

有关此分析的外观和功能的信息，请参阅按身份构件列出的单一身份配置文件。

未分段的轮廓 unsegmented-profiles

通过此洞察回答的问题：

  • 有多少用户档案不属于某个受众？
  • 未分段配置文件占总受众的百分比是多少？
  • 任何合并策略是否都会产生大量未分段的配置文件？
选择以显示生成此分析的厂蚕尝
code language-sql 
SELECT qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_name,
       sum(qsaccel.profile_agg.adwh_fact_profile.count_of_Orphan_profiles) CNT
  FROM qsaccel.profile_agg.adwh_fact_profile
  LEFT OUTER JOIN qsaccel.profile_agg.adwh_dim_merge_policies ON qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_id=adwh_fact_profile.merge_policy_id
  WHERE qsaccel.profile_agg.adwh_fact_profile.date_key='2024-01-10'
    AND qsaccel.profile_agg.adwh_fact_profile.merge_policy_id = 2027892989
  GROUP BY qsaccel.profile_agg.adwh_dim_merge_policies.merge_policy_name;

有关此分析的外观和功能的信息，请参阅未分段配置文件小组件文档。

后续步骤

通过阅读本文档，您现在了解了生成仪表板分析的SQL以及此分析可以解决哪些常见问题。 您现在可以对SQL进行编辑和迭代，以生成您自己的见解。

有关如何直接通过PLatform UI调整分析的SQL的详细信息，请参阅查看厂蚕尝文档。

您还可以阅读并了解为受众、帐户配置文件和目标仪表板生成分析的厂蚕尝。

recommendation-more-help
ececc77d-ff44-4382-85ee-a087c8834323