Operating scope

Country head

Capability

All capabilities

Active handovers

Outcomes written back

Connected work · Agent system

Business vertical

People and workforce

Agent on shift

Screening Agent

Accountable person

Hiring managers

Evidence · hourly

Roster and attendance

Next step

PeopleOS

PeopleOS · screen 03 of 05

First ninety days

Most frontline leavers go in the first three months, and almost all of it is visible beforehand.

3 issues · 2 waiting on you

Issues in scope

11 instances across 11 places

Waiting on you

0 already decided in this session

Three largest

AED 2.9m–AED 9.8m

Modelled annual value, three largest only

Running without asking

0 armed and waiting on a threshold

Decision owner

Country head

Returns to Country head, Group

Scope

Oman

27 stores · all capabilities

Listen

Start date and storeShift pattern actually worked against what was offeredTraining completionEarly absence pattern

Process

Is the job what was describedIs the shift pattern what was agreedWhich stores lose people fastest

Act

Check-in conversation in their languageSurface the store pattern, not the personHand anything material to HR

Written backEvery outcome returns to the systems that produced the evidence.

KindOrder

Queue

3 issues across 11 places · 2 waiting on you

Approve

26% of frontline joiners in Oman leave inside ninety days

Oman

What happened

The leavers are concentrated in specific stores rather than spread across the estate, and the shift pattern actually worked in those stores differs from what was described at offer more often than elsewhere.

Why it matters

Every one of these was recruited, screened, cleared, trained and rostered. Replacing them costs the whole cycle again plus the coverage gap while the role is open. Concentration in specific stores means this is a management pattern, not a recruitment failure.

Already done

Early leavers are analysed at store level with the offered versus worked shift pattern attached, rather than reported as a group attrition rate.

Worth

AED 2.8m to AED 9.5m a yearModelledFrontline joiners leaving inside ninety days after the full recruit, clear, train and roster cycle. Value is the replacement cycle avoided at the concentration observed.Ninety day frontline attrition

Then what

Store level pattern with the shift evidenceCountry headReports back 30 Sept

If approved

The store level pattern goes to the Oman country head with the shift evidence

If declined

It stays a group attrition percentage, which is a number nobody can act on

Draft

New joiners at Sohar are not asked anything until they leave

Sohar

What happened

The first structured conversation with a frontline joiner is usually the exit interview. Nothing happens at week two, when the answer would still be useful.

Why it matters

A two minute check-in in the joiner's own language, at week two, catches the shift mismatch, the training gap and the manager problem while all three are still fixable. After week eight the person has usually already decided.

Already done

A week two check-in is drafted, in the languages the store's joiners actually speak.

Worth

AED 94k to AED 309k a yearModelledA two minute check-in at week two catches the shift mismatch while it is still fixable. Value is the share of early leavers recoverable at that point.Joiners asked anything before exit

Then what

Week two check-in sent to current joinersCountry HRReports back 19 Sept

You joined Lulu Sohar two weeks ago. Three quick questions: are the shifts what you expected, have you had the training you need, and is there anything getting in your way? Your answers go to HR, not to your store manager.

Also open at

Declined

Check-in answers are reported as store patterns, never as named individuals

Ibra

What happened

The check-in produces material about specific managers and specific stores. It is aggregated before anyone sees it, and no individual response is attributable.

Why it matters

A check-in that a joiner believes is attributable produces nothing true. The only version worth running is one where the person can say the shifts are wrong without it arriving on their manager's desk with their name on it.

Already done

Aggregated at store level with a minimum response threshold before anything is reported at all.

Exposure

Not a revenue lineAn attributable check-in produces answers that are polite and useless, which is worse than not asking because it looks like evidence.Check-ins reported at store level only

Then what

Refusal held, store level aggregate onlyGroupStanding, no return date

Override

Nobody. Refused at design

Cost of overriding

An attributable check-in produces answers that are polite, useless, and worse than asking nothing