19804091 CustomerApi.Jobs.Insights.ExecuteAccountInsightAgentJob
Queue insight_agents
Attempt 1 of 2
Priority 0
Tags ...
Node customer_api@10.10.1.109
Queue Time 00:00.014
Run Time 00:53.376
Inserted 3h ago
Scheduled 3h ago
Completed 3h ago (00:54)
Cancelled —
Discarded —

Args 

%{
  "account_id" => "106160",
  "account_insight_agent_id" => "62e3b1cf-db85-43c3-82f1-0205b271a4b3"
}

Meta 

%{
  "deps" => [],
  "name" => "execute",
  "on_hold" => false,
  "partition" => true,
  "partition_key" => "Sh7hncd3qVugngnnlRyQKGtNgFvW1lu8JNjBZhmMs7g",
  "recorded" => true,
  "return" => "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",
  "structured" => true,
  "uniq" => true,
  "uniq_bmp" => [1, 0, 2, 3],
  "uniq_key" => "J7r5pkkoxWVm4n10DvbwNVQNiugvrIzkcB+hm1fhIN8",
  "workflow" => true,
  "workflow_id" => "019cbfb1-6933-763c-988c-ee411607fd3b"
}

Recorded Output 

%{
  insights_data: [
    %{
      "action_plan" => "1) Confirm Appcues tracking is active (SDK installed and identifying users consistently).\n2) Ensure you have a reliable “first login” anchor (either a **first_login** event or a user property like **first_login_at / signup_at**).\n3) Once tracking is live, re-run the analysis in Appcues using: a cohort of new users (first login window), the “3 banner impressions within 14 days” condition, and a Month-2 retention proxy (e.g., any session/page view/custom “active” event in days 31–60).",
      "cta_text" => nil,
      "entity_id" => nil,
      "entity_type" => nil,
      "expected_impact" => "Unlocks the ability to verify whether the uplift is real (vs. selection bias) and then optimize delivery to increase adoption above 1.04%.",
      "finding" => "Your Loops insight is directionally interesting (1.04% adoption of a 3-banner/14-day cadence correlates with **+146.4% Month-2 retention** for adopters), but Appcues currently shows **0 tracked users, 0 events, and 0 banners/messages delivered** in the last 30 days. With no in-product tracking, we can’t confirm the cohort definition, adoption rate, or retention lift inside Appcues—or use Appcues to iterate on it.",
      "outcomes" => ["retention", "adoption"],
      "priority" => "high",
      "title" => "Can’t validate the Loops banner/retention uplift in Appcues yet (no tracked product data in the last 30 days)",
      "type" => "alert"
    },
    %{
      "action_plan" => "Create a simple experiment design:\n1) Define an eligible population: all users within 0–14 days of first login.\n2) Randomly split into two groups: **Cadence ON** vs **Holdout** (no banners or a neutral control).\n3) Measure retention the same way Loops did (or with a clear in-app “active in Month 2” definition) and compare Month-2 retention between groups.\n4) Only scale if the holdout-adjusted lift remains meaningful.",
      "cta_text" => nil,
      "entity_id" => nil,
      "entity_type" => nil,
      "expected_impact" => "Prevents over-investing in a tactic that only appears to work because the most engaged users are the ones who adopt it; increases confidence in retention impact.",
      "finding" => "At **1.04% adoption**, the “adopting group” is usually a tiny, highly engaged subset. That often produces big-looking retention lifts even if the banners aren’t the cause (classic selection bias). Without a randomized comparison group that was eligible but *didn’t* receive the cadence, it’s not possible to attribute the uplift to the banners with confidence.",
      "outcomes" => ["retention"],
      "priority" => "high",
      "title" => "The 146% retention lift is likely inflated by self-selection at 1.04% adoption—run a holdout test before scaling",
      "type" => "risk"
    },
    %{
      "action_plan" => "Once tracking is validated, derive these additional insights and act on them:\n1) Delivery diagnostics: for each of the 3 banners, measure **seen rate** and **dismiss rate** among new users.\n2) Sequence completion: how many users see **1 → 2 → 3** within 14 days? Identify the step with the biggest drop.\n3) Targeting refinement: segment by key cohorts (role, plan tier, acquisition channel, platform) to find where the cadence works best.\n4) Content iteration: A/B test the first banner’s hook and CTA (it’s the “gateway” to adoption).",
      "cta_text" => nil,
      "entity_id" => nil,
      "entity_type" => nil,
      "expected_impact" => "If you can move adoption from ~1% to even **3–5%** while preserving some of the retention lift, the overall retention impact becomes materially larger.",
      "finding" => "Even if the Loops relationship holds, the limiting factor is reach: only **~1 in 100 users** “adopt” the 3-banner sequence. That suggests the cadence is either not being delivered broadly to new users, is easy to dismiss/ignore, or is being shown at the wrong time (too early, too late, or during low-intent moments).",
      "outcomes" => ["adoption", "retention", "engagement"],
      "priority" => "medium",
      "title" => "Biggest immediate upside: raise adoption above 1.04% by improving who sees the 3-banner cadence and when",
      "type" => "opportunity"
    }
  ]
}

Errors 

No Errors