CSM team · Grok pilot briefing

Grok is quietly saving CSMs hours every week

24 repeat tasks logged by 5 CSMs. The first 8 measured cases cut time per run by 68%.

Prepared for Chris · Source: CSM Success Case and Agent Build Log, 12 Oct 2026

21 of 24 logged use cases are already built

repeat tasks logged

built and complete (88%)

validated in testing (50%)

CSMs contributing

Use cases logged per CSM

Dave

Zaira

Linda

Nixie

Melvin

Source: Success Cases tab, rows SC-001 to SC-024. Build status Complete = 21, Ready 2, Building 1.

194 hours a month of repeat work is in scope

Baseline time before Grok, from volume per month times minutes per run. Five tasks hold 107 of those hours (55%).

Morning check: calendar, invoices, QBR follow-ups

40.3 hrs

Weekday inbox triage + draft replies

33.0 hrs

Vital Signs scoring prep + risk flags

12.0 hrs

Coaching session prep (day ahead)

11.3 hrs

WhatsApp scheduling messages to calendar

10.0 hrs

Source: Baseline monthly hours, Success Cases tab. The other 19 cases total 88 hrs.

Measured: 82 hours a month saved in 8 cases

68%

average time saved per run across the 8 measured cases

116 → 34

hours a month, before and after, for those 8 cases

Task	Before	After	Saved
Morning check + collections draft	110 min	30 min	73%
Inbox triage + draft replies	90 min	30 min	67%
AD coverage vs payroll check	90 min	11 min	88%
Monthly client update emails	30 min	15 min	50%
Quarterly QBR emails + nudges	30 min	15 min	50%
Coaching prep + post-notes	30 min	11 min	63%
Vital Signs scoring prep	180 min	30 min	83%
WhatsApp scheduling to calendar	30 min	8 min	73%

Source: SC-010 to SC-017. Minutes per run. After = agent run plus the CSM's review, as estimated by the CSM.

Savings come from daily routines and heavy prep

Daily routines

51 of the 82 hours

The morning check and weekday inbox triage run about 22 times a month each. Small cuts per run add up fast.

Heavy prep

3 hours down to 30 minutes

Vital Signs scoring prep fell from 180 to 30 minutes. AD payroll reconciliation fell from 90 to 11.

CSM stays in control

Review time is counted

The after figures include the CSM's own check, 5 to 7 minutes in the notes. The saving is net of that.

Source: Success Cases tab, SC-010 to SC-017, Notes column.

If the other 16 cases match, savings reach about 136 hours a month

Measured

82 hrs

saved a month in the 8 cases with actual times

Projected

54 hrs

more, if the other 16 cases (79 baseline hrs) save 68%

Total if realized

136 hrs

a month, about 70% of the 194 baseline hours

The projection is an assumption, not a result. The measured cases are the largest tasks, so lighter tasks may save less in hours.

Source: Success Cases tab. Projection = 78.7 baseline hrs × 68.4% average measured saving.

Four things to close before this goes wider

Measure the other 16 cases

Only 8 of 24 have an Actual min/run. Nixie is asking each CSM to fill it in.

Confirm client-data approvals

16 of 24 cases say client data is used. The Summary tab flags 8 rows without recorded approval.

Finish quality testing

13 cases pass, 8 pass with revisions, 3 are not tested yet (SC-013, SC-015, SC-017).

Track adoption and training

The log has no usage or training data. Add weekly active CSMs and sessions attended.

Source: Success Cases tab and Summary tab, 12 Oct 2026.