Accounts at risk
Counts accounts in predefined engagement-decline segments and tracks the weekly total. Uses fixed segments as an approximate at-risk signal.
Authored by Intempt Technologies
Recipe curators
Agentic workflow
Coming soon1 step Blu runs end to end. It drafts a plan and nothing runs until you approve it.
- 1
Count accounts going quiet
Coming soonBlu builds a step
About this recipe
When to use it
By the time an account churns, it's too late to save it. This tracks the count of accounts whose engagement has already dropped materially, so CS gets the warning while there's still time to intervene.
What it delivers
Counts and trends accounts whose engagement has declined past a material threshold, giving customer success one headline early-warning number instead of digging through individual account histories.
How you prompt this
Describe the outcome in your own words and Blu Agent maps it to this recipe, or run it by name. It drafts a plan first and nothing runs until you approve it.
Tell Blu Agent what you want
“Counts accounts in predefined engagement-decline segments and tracks the weekly total. Uses fixed segments as an approximate at-risk signal.”
recipe.md preview
Built from the published catalog. Blu Agent's full step prompts load in your workspace when you run it.
Inputs & outputs
What the recipe needs before it runs, and what it leaves in your workspace.
Inputs
- Nothing declared beyond an Intempt workspace. Blu asks for anything else it needs as it goes.
Outputs
How the recipe runs
Blu follows the steps in order and shows you each one before anything goes live.
- 01Count accounts going quietCreate an Insights report called "Accounts at Risk". Definition of "at risk": An account (Users.the account the user belongs to) qualifies as at-risk if ALL of the following hold for the trailing 4 weeks vs. the prior 8-week baseline: - Distinct active users in the account dropped by ≥30% (count of unique users from the account who emitted a Session start) - OR the dominant user (most-active user on the account) has emitted zero Session start events in the last 14 days (champion-departure proxy) - OR the account's primary_user_id has emitted ≥1 Ticket created with priority = high or P1 in the last 30 days Series A: Count Unique values of the account the user belongs to satisfying the at-risk criteria, time granularity: Weekly snapshot, label: "Accounts at Risk" Series B: Count Unique values of the account the user belongs to NOT satisfying the criteria but whose engagement dropped 10: 30% (early-warning band), label: "Accounts to Watch" Series C: Sum of subscription amount (most-recent active Subscription started.amount) across at-risk accounts: the dollar revenue at risk, unit: $, label: "Revenue at Risk" Time range: Last 12 weeks Breakdown: By the account the user belongs to segment if the project has account-tier metadata (Enterprise / Mid-Market / SMB) on the Accounts object; otherwise by trailing-30d engagement-volume bucket Compare: Previous period (prior 12 weeks) Chart type: Stacked area chart for Series A and B over time, with Series C ($) as a secondary line on a right axis Annotations: - Flag any week where Series A grew by ≥20% vs. the trailing 4-week average (acute risk spike: investigate what happened that week). - Surface the top 5 accounts currently in Series A by subscription amount (highest-value at-risk accounts get priority CSM action). - Flag if Series A as % of total active accounts exceeds 15% (broad-base health issue, not isolated account problems). - Highlight the dominant trigger: which of the three at-risk criteria fires most often? If "distinct users dropped" dominates, the issue is account-level disengagement; if "champion departure" dominates, the issue is single-threading; if "high-priority tickets" dominates, the issue is product quality. Use case: the canonical CS headline metric. Most CS teams maintain this list manually in a spreadsheet; this recipe automates it. Distinct from account-engagement-score-trend (which is the per-account engagement timeline): this is the count headline. Taxonomy notes: - Session start, Ticket created are canonical events. Users.the account the user belongs to is the canonical account-level relation. - "Champion" identification = the user with the highest count of Session start events on the account in the trailing 30 days. "Departure" = zero Session start events from that user in the last 14 days. - Ticket created.priority is a real property; the "high-priority" classification depends on the project's priority enum (typically P1, P2, P3, P4 or High/Medium/Low). - Subscription amount lookup: the account's primary_user_id's most-recent active Subscription started.amount, summed across users on the account if the workspace has multi-subscription accounts.
Install Intempt and run it
- 1
Sign up
Create a free Intempt workspace.
- 2
Connect your data
Add the sources this recipe reads from.
- 3
Ask Blu
Type /accounts-at-risk-count in Blu once this recipe is live there.
- 4
Review and approve
Blu builds each step and waits for your OK before anything goes live.
