UAE · ENBook a demo
Procure AICustomerInvestigator

Client Account Health Agent

Scores accounts on usage and sentiment.

Client Account Health AgentAccount Health Agent · Customer Lifecycle
Industry
Procure AI
Department
Customer
Team
Customer Lifecycle
Kind
Investigator
Approver
Client Engagement Lead
Code
CUS-13

It scores each client's health from platform adoption (active requisitioners, share of spend through POs), service results (realised savings against target, first-time-match rate), support friction, survey sentiment and fee billing, using the weighted health model. It explains which signals moved the score and the likeliest cause, writes the score to the CRM and gives the engagement lead a task when a client turns red. **The problem it solves:** Why each client's health score is what it is and what moved it: the adoption, results, support, sentiment and billing signals over the period, and the events that could explain them.

How it works

6 steps, and a person signs off before step 6.

  1. 01Looks it up

    Finds the record the request is about in the spend analytics warehouse: the adoption, PO coverage, savings, first-time-match, ticket, survey and fee billing figures for the clients over the period..

  2. 02Reads the request

    Works out what is being asked and what a complete answer needs.

  3. 03Checks the facts

    Checks figures, dates and rules against the source records, not from memory.

  4. 04Works it through

    Weighs competing explanations against the evidence and shows which one holds up.

  5. 05Asks a person

    Hands the work to Client Engagement Lead with its reasoning before anything goes out.

  6. 06Acts

    Once approved, it acts in the CRM: it updates each client's health score, band and drivers, and creates a task for the engagement lead when it turns red. and records what was done.

What it works with

Your knowledge, your systems, your people.

Knows

The client health model: signal definitions, weights, the green, amber and red thresholds, KPI targets per service scope and the playbook step for each band.

Reads from

the spend analytics warehouse: the adoption, PO coverage, savings, first-time-match, ticket, survey and fee billing figures for the clients over the period.

Acts in

the CRM: it updates each client's health score, band and drivers, and creates a task for the engagement lead when it turns red.

Hands back

A list with one entry per client, each with client_account, score (0 to 100), band (green, amber or red), change (points since the previous score, or null), drivers (a list, each with signal, value and effect), likely_cause, how_to_confirm and next_step.

Asks a person when

A client turns red, its score drops 15 points or more, or it is scored on fewer than three signals.

ApproverClient Engagement Lead

Settings

Your limits, set by your managers.

SettingDefaultWhat it controls
Weight of adoption (%)30Share of the score from active requisitioners and share of spend through POs.
Weight of service results (%)30Share of the score from savings against target and first-time-match rate.
Weight of support (%)15Share of the score from ticket volume and severity.
Weight of sentiment (%)15Share of the score from NPS and QBR feedback.
Weight of billing (%)10Share of the score from overdue fee invoices.
Red below50A score below this is red; below 75 and not red is amber.
Points drop that needs an explanation15A score change of this many points or more since the last score is explained and raised.

Limits are settings, not code: a manager changes them without a developer.

Rules it always follows

  • Score each client from 0 to 100 as adoption 30, results 30, support 15, sentiment 15 and billing 10 percent; if the weights do not add to 100, scale them and say so.
  • Band the score red below 50, amber below 75 and green otherwise.
  • Use only finance-validated savings for the results signal, never pipeline or cost avoidance counted as hard savings.
  • Leave out a signal with no data and rescale the other weights, and never score a client on fewer than two signals.
  • Explain every change of 15 points or more by the signal that moved it and the likeliest cause, and say what would confirm it.
  • Prefer an explanation backed by a dated event, such as an ERP migration or a sponsor leaving, over a general trend.

See the Client Account Health Agent agent on your own work.

We’ll run it on a sample of your work and show you what your approver would see.

Client Account Health Agent · Procure AI | Basinwright Industries