A Number Is Not a Decision

When I took on the Operational Resource Management role for a service line in a global technology services organisation, the bench stood at 235 people across India, the Philippines, Poland and Portugal.

Nobody was short of data. Bench data lived in the resource management system, and each global delivery centre kept its own Excel report of its bench position. What leadership lacked was the visibility to act on that information, and to decide what should happen to each person on the bench.

That gap is easy to mistake for a reporting problem. Fix it as one and you get a better report and the same bench. A number moves only when someone decides, the decision is recorded, and someone carries it out.

Data is not visibility

A bench is not a static number. Several pressures act on it at once: the financial pressure of people waiting without billable work, how long each person has been waiting, unpredictable releases from projects, including project and customer exits, redeployment into new work, and the match between demand and available skills.

A report kept separately by each delivery centre describes that centre's position well. It does not, by itself, tell a leadership team where to act across four countries, who should act, or by when.

From number to action

What we put in place was a chain with five links. Each depends on the one before it.

1. Data

   The resource management system remained the source of bench data. Each delivery centre's Excel report remained its local view.

2. ↓

   Shared visibility

   We brought the existing information into one central ORM governance repository. The ORM team maintained it across all service lines and delivery centres, and it became the single source of truth for reconciling bench data.

3. ↓

   Decision

   L2 and L3 leadership recorded a decision on each person on the bench: retain, redeploy or terminate. The Bench High Ageing tool, developed in Power Apps, supported the decision workflow.

4. ↓

   Review

   Every week I led a governance review across all capability lines in the service line. The shared view and the recorded decisions were reviewed there.

5. ↓

   Follow-through

   The ORM team acted on the recorded decisions.

Consolidating the data covers only the first two links. In my view, that is where many organisations stop. The chain's value lies in connecting shared visibility to recorded decisions, regular review and follow-through.

What a recorded decision led to

Each decision category had its own route to action:

• Terminate. Termination actions were sent to HR for action.
• Redeploy. Redeployment action was taken where new project demand and suitable opportunities were visible.
• Retain. People were retained where there was a strategic need for niche skills, or new pipeline was anticipated.

None of these was instant. Redeployment depended on suitable demand. Retention depended on a genuine skill need or a visible pipeline. The repository and governance process gave recorded decisions a route to action, rather than leaving them as entries in a record.

The weekly review

The weekly review covered five areas, and each was a decision rather than an update:

• Reviewing bench ageing.
• Assigning redeployment actions.
• Resolving mismatches between demand and capacity.
• Escalating exceptions.
• Agreeing financial actions.

The cadence created accountability and made decisions timely. Alongside it sat a global bench policy, and we acted on it with discipline across all four countries. A policy is easy to write. Applying it every week, in every centre, is where it earns its value.

The result, and what it does not prove

Over seven months the bench fell from 235 to 23 across India, the Philippines, Poland and Portugal: 212 people, about 90%.

I would not credit that to any single intervention. It came from the whole chain working together, and from the redeployment and follow-through work done in each delivery centre.

Controlled, not predictive

We moved bench management from reactive to controlled. We did not reach predictive. Predictive demand modelling and preventive workforce planning were the next stage, identified but not achieved.

Looking back, I would establish the shared decision record and governance cadence earlier, rather than leave reporting distributed across local spreadsheets. I would also bring demand visibility and release forecasts into the review sooner, creating a stronger basis for anticipating future capacity pressures.

The question to ask

If your data is sound and the number still does not move, follow the chain. Who can see the data and decide on it? Is each decision recorded? Who reviews it, and how often? And who carries it out?