From 9af498833e381d48d8c5b371ec6e5659612cde52 Mon Sep 17 00:00:00 2001
From: Karl-Aksel Puulmann <macobo@users.noreply.github.com>
Date: Tue, 19 Nov 2024 10:12:39 +0200
Subject: [PATCH] Channels: backfill utm_medium based on click_param_id (#4833)

* Backfill utm_medium

Follow-up to https://github.com/plausible/analytics/pull/4817

* Update backfill
---
 CHANGELOG.md                                  |  1 +
 ...238_backfill_utm_medium_click_id_param.exs | 27 +++++++++++++++++++
 2 files changed, 28 insertions(+)
 create mode 100644 priv/ingest_repo/migrations/20241118112238_backfill_utm_medium_click_id_param.exs

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 7090b49ff..dc0f3056f 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -5,6 +5,7 @@ All notable changes to this project will be documented in this file.
 
 ### Added
 - Dashboard shows comparisons for all reports
+- UTM Medium report and API shows (gclid) and (msclkid) for paid searches when no explicit utm medium present.
 
 ### Removed
 ### Changed
diff --git a/priv/ingest_repo/migrations/20241118112238_backfill_utm_medium_click_id_param.exs b/priv/ingest_repo/migrations/20241118112238_backfill_utm_medium_click_id_param.exs
new file mode 100644
index 000000000..df0c470a2
--- /dev/null
+++ b/priv/ingest_repo/migrations/20241118112238_backfill_utm_medium_click_id_param.exs
@@ -0,0 +1,27 @@
+defmodule Plausible.IngestRepo.Migrations.BackfillUtmMediumClickIdParam do
+  @moduledoc """
+  Backfills utm_medium based on referrer_source and click_id_param
+  """
+  use Ecto.Migration
+
+  def up do
+    execute(fn -> repo().query!(update_query("events_v2")) end)
+    execute(fn -> repo().query!(update_query("sessions_v2")) end)
+  end
+
+  def down do
+    raise "irreversible"
+  end
+
+  defp update_query(table) do
+    """
+    ALTER TABLE #{table}
+    UPDATE utm_medium = multiIf(
+      referrer_source = 'Google' AND click_id_param = 'gclid', '(gclid)',
+      referrer_source = 'Bing' AND click_id_param = 'msclkid', '(msclkid)',
+      utm_medium
+    )
+    WHERE empty(utm_medium) AND NOT empty(click_id_param)
+    """
+  end
+end