Backlog Is Growing — Do You Need More People or a Better Process?

Backlog is growing.

SLA is deteriorating.

Employees say they are overloaded.

The most obvious conclusion is:

“We need more people.”

Sometimes that conclusion is correct.

But not always.

A backlog can increase because the operation genuinely lacks capacity.

It can also increase because available capacity is being used inefficiently.

For example:

• Incoming volume increased.
• AHT increased.
• Employees are working the wrong queues.
• Only a few people have the required skills.
• Cases are being reworked.
• Follow-ups are not returning to the right queue.
• Difficult work is being avoided.
• A new process created additional steps.
• Productivity assumptions are unrealistic.
• Capacity is available, but not when or where demand occurs.

Hiring may solve some of these problems.

But if the real issue is process design, adding headcount can simply make an inefficient process larger.

The better question is:

What is actually causing the backlog to grow?

A useful diagnostic sequence is:

Demand

↓

Available Capacity

↓

Productivity

↓

Work Mix

↓

Skills

↓

Work Allocation

↓

Process Bottlenecks

↓

Quality and Rework

↓

Backlog and SLA Impact

Only after understanding these drivers should management decide whether the answer is additional staffing, process improvement, or a combination of both.

Backlog Is a Mathematical Outcome

At its simplest:

Closing Backlog = Opening Backlog + Incoming Work − Completed Work

Suppose:

Opening backlog = 10,000 cases

Incoming volume = 2,000

Completed volume = 1,700

Closing backlog:

10,000 + 2,000 − 1,700 = 10,300

Backlog increased by:

300 cases

That tells us what happened.

It does not tell us why.

The next step is understanding why completions were lower than required.

First Question: Did Demand Increase?

Suppose the team normally receives:

10,000 cases per week

but this week receives:

13,000

That is a:

30% increase in incoming demand

If workforce capacity remains unchanged, backlog may increase even if employee productivity remains perfectly stable.

For example:

Normal capacity = 11,000 cases

Normal incoming demand = 10,000

Spare capacity = 1,000

Now demand increases to:

13,000

Capacity remains:

11,000

Weekly shortage:

2,000 cases

Backlog growth is therefore expected.

This is primarily a capacity-vs-demand problem.

The operation may genuinely need:

• Additional staffing
• Overtime
• Temporary resources
• Automation
• Higher productivity
• Demand reduction
• A revised SLA

depending on how permanent the demand increase is.

Temporary Volume Spike or Structural Growth?

The nature of the increase matters.

Suppose volume increased for one week because of a seasonal event.

Hiring permanent employees may not be the best solution.

Now suppose volume has grown:

January — 50,000

February — 54,000

March — 58,000

April — 62,000

The pattern suggests structural growth.

In that situation, a workforce-capacity review may be necessary.

This is why demand should be analyzed as:

Actual vs Forecast

and:

Temporary vs Sustained

before hiring decisions are made.

Second Question: Did Capacity Fall?

Backlog can increase even when demand remains stable.

Suppose:

Incoming volume = 50,000

in both March and April.

March available productive capacity:

52,000 cases

April available productive capacity:

45,000 cases

Now the April operation has:

5,000 cases of capacity shortage

without any increase in incoming demand.

Possible causes include:

• Higher absence
• Approved leave
• Attrition
• Training
• New-hire ramp-up
• Reduced productive hours
• System downtime
• Employees transferred to other work

The problem is not higher demand.

It is lower available capacity.

Gross Headcount Can Hide a Capacity Drop

Suppose gross headcount remains:

100 employees

Management may assume workforce capacity is unchanged.

But April includes:

5 employees on leave

5 employees in training

10 new hires at 50% productivity

5 employees assigned to another activity

The effective workforce is materially lower than 100 fully productive FTE.

This is why backlog analysis should use:

Effective Productive FTE

rather than relying only on payroll headcount.

For more detail, read Why Your Operations Team Can Be Understaffed Even When Headcount Looks Sufficient.

Third Question: Did AHT Increase?

Sometimes headcount and demand are stable, but each transaction takes longer.

Suppose monthly volume is:

60,000 cases

Average Handling Time increases from:

10 minutes

to:

12 minutes

At 10 minutes:

Required handling time:

60,000 × 10 ÷ 60 = 10,000 hours

At 12 minutes:

60,000 × 12 ÷ 60 = 12,000 hours

The process now requires:

2,000 additional handling hours

without any change in volume.

If one productive FTE contributes:

154 productive hours per month

the additional requirement is approximately:

2,000 ÷ 154 = 13 productive FTE

That is a major capacity change.

Before hiring 13 people, management should ask:

Why did AHT increase?

Why Can AHT Increase?

Possible reasons include:

Work Mix

More complex transactions.

Process Change

Additional validation or documentation steps.

System Performance

Slow applications or increased downtime.

New Employees

Higher proportion of employees still learning.

Policy Change

Additional required checks.

Poor Upstream Quality

Cases require more investigation.

Workflow Design

Employees are switching between multiple systems or queues.

Some causes may require more people.

Others may be solved more effectively through process improvement.

Fourth Question: Did the Work Mix Change?

Volume alone does not represent workload.

Suppose monthly volume remains:

50,000 transactions

But the mix changes.

Previous Month

35,000 simple

10,000 standard

5,000 complex

Current Month

20,000 simple

15,000 standard

15,000 complex

Assume:

Simple = 6 minutes

Standard = 12 minutes

Complex = 30 minutes

Previous workload:

8,000 handling hours

Current workload:

12,500 handling hours

Volume is unchanged.

Workload has increased by:

56.25%

Hiring based only on transaction count may therefore produce the wrong conclusion.

A capacity model should consider the amount of work effort, not only the number of work items.

Fifth Question: Is the Problem Concentrated in One Queue?

Suppose total backlog increased:

15%

Management may see a department-wide problem.

But drill-down shows:

The backlog problem is overwhelmingly concentrated in:

Queue C

Now the question becomes:

What is happening in Queue C?

Possible explanations include:

• Volume increased
• AHT increased
• Skill capacity is insufficient
• Routing rules are ineffective
• A particular case type is blocked
• Follow-ups are returning faster than expected
• Rework increased

A department-level staffing decision may be unnecessary.

The problem may require intervention in one queue.

Sixth Question: Do You Have a Skill-Capacity Problem?

Suppose:

Total effective FTE = 60

Required FTE = 58

Overall capacity appears sufficient.

But:

Overall:

Available = 60

Required = 58

The operation appears:

+2 FTE overstaffed

Yet Complex Processing is:

-6 FTE understaffed

If the growing backlog is primarily Complex Processing, hiring more standard processors may achieve nothing.

The solution could instead be:

Cross-training

Skill reassignment

Specialist hiring

or:

Process simplification

This is why workforce planning must sometimes operate at skill level rather than department level.

Seventh Question: Is Work Being Allocated Properly?

Capacity can exist and still fail to reach the cases that need it.

Imagine employees can manually choose whichever case they want.

Simple work may be completed quickly.

Difficult work remains.

Total production looks respectable.

But old, complex cases accumulate.

Backlog aging deteriorates.

Management may interpret this as a staffing shortage.

The underlying problem may be work selection.

A more structured allocation model might consider:

Skill Eligibility

↓

Priority

↓

SLA Risk

↓

FIFO

This helps make sure capacity is directed toward the appropriate work rather than relying entirely on employee selection.

For more on this topic, read FIFO vs Priority vs Skill-Based Work Allocation.

Cherry-Picking Can Distort Capacity Reporting

Suppose two employees each work eight productive hours.

Employee A completes:

48 easy cases

Employee B completes:

24 complex cases

Raw CPH:

Employee A = 6

Employee B = 3

If the complex cases require twice as much handling time, their performance may actually be comparable.

If employees are allowed to consistently choose easier work, total output may appear healthy while difficult cases age.

That can create:

High overall production

and:

Growing operational backlog

at the same time.

This is why work mix and allocation should be investigated before using raw output to justify staffing decisions.

Eighth Question: Is Rework Consuming Capacity?

Suppose:

Incoming external volume = 50,000

The team completes:

50,000

Management concludes:

Capacity equals demand.

But:

5,000 cases require rework

Now the team effectively has another 5,000 processing events to handle.

The real workload is closer to:

55,000 units of processing activity

depending on how much effort each rework item requires.

Rework can therefore create a hidden capacity requirement.

A backlog problem may actually be a quality problem.

Example: Quality-Driven Capacity Loss

Suppose:

5,000 rework cases

Average rework handling time = 8 minutes

Additional workload:

5,000 × 8 = 40,000 minutes

or:

666.7 hours

If one productive FTE provides:

154 hours per month

rework consumes approximately:

666.7 ÷ 154 = 4.3 FTE

The business could potentially recover more than four FTE of capacity by reducing the underlying defects.

Hiring four or five additional employees would increase capacity.

Fixing the quality problem might remove the need for that additional capacity altogether.

Ninth Question: Are Follow-Ups Creating Hidden Work?

Backlog may contain more than new work.

Cases may move through:

Process

↓

Pend

↓

Follow-Up

↓

Process Again

↓

Complete

If follow-ups are poorly controlled, cases can:

• Remain pended too long
• Become overdue
• Return in large waves
• Require repeated handling
• Lose clear ownership

This generates additional workload that may not appear in simple received-volume reporting.

Management should therefore distinguish between:

New Cases

and:

Existing Cases Requiring Additional Activity

For more detail, read How to Manage Follow-Ups Without Excel and Outlook Reminders.

Tenth Question: Is One Process Step the Real Bottleneck?

Consider a workflow:

Intake

↓

Validation

↓

Processing

↓

Review

↓

Completion

Suppose:

Intake capacity = 1,000/day

Validation = 1,000/day

Processing = 1,000/day

Review = 600/day

Completion = 1,000/day

The process can only sustainably deliver around:

600 cases per day

because Review is the bottleneck.

Adding staff to Intake does not improve end-to-end throughput.

It may simply create a larger queue waiting for Review.

Management needs to identify the constrained step before increasing headcount.

Bottleneck Capacity Determines Flow

This principle is important.

If every process stage can handle 1,000 cases except one stage that can handle 600:

End-to-end sustainable throughput cannot reliably exceed the constrained stage without changing the process or adding capacity there.

This may require:

• Additional reviewer capacity
• Cross-training
• Process simplification
• Automation
• Removing unnecessary review
• Changing risk thresholds

The answer depends on why the bottleneck exists.

Eleventh Question: Is the Process Creating Unnecessary Work?

Operations can accumulate process steps over time.

For example:

An employee:

Downloads a report.

Copies data into Excel.

Checks another application.

Updates a tracker.

Sends an email.

Updates the original system.

Each action may have been added for a good reason.

But collectively, the process may contain significant manual handling.

If employees spend:

20% of their productive day

performing avoidable administrative activity, hiring more people increases total capacity but preserves the inefficiency.

Process improvement may recover capacity from the existing workforce.

Capacity Released Through Process Improvement

Suppose:

50 employees

7 productive hours each

Current process wastes:

30 minutes per employee per day

Total recoverable time:

50 × 0.5 = 25 hours per day

If CPH = 5

potential theoretical additional capacity:

25 × 5 = 125 cases per day

Over 22 working days:

125 × 22 = 2,750 cases per month

Even relatively small process improvements can therefore create meaningful capacity at scale.

Test whether more headcount will actually solve the backlog

Before adding people, model what happens if you change staffing, AHT/CPH, shrinkage, automation or working days.

The free Praevexa Workforce Planner lets you compare capacity against incoming workload and backlog so you can see whether the real problem is insufficient HC, low productivity, or a process constraint.

Test Your Backlog and Staffing Plan

Link to:

https://www.praevexa.com/WorkforcePlanner.aspx

Twelfth Question: Are Productivity Targets Realistic?

Suppose workforce planning assumes:

6 CPH

Actual stable performance:

5 CPH

Available productive hours:

10,000

Planned capacity:

10,000 × 6 = 60,000 cases

Actual capacity:

10,000 × 5 = 50,000 cases

Demand:

55,000

The planning model says:

5,000 spare capacity

Reality suggests:

5,000 capacity shortage

The backlog is not necessarily caused by employees suddenly underperforming.

The planning assumption may simply be unrealistic.

Before deciding whether to hire or manage performance, validate the target itself.

Thirteenth Question: Is Productivity Actually the Problem?

Suppose:

Backlog = +20%

CPH = stable

Quality = stable

AHT = stable

Incoming volume = +22%

The evidence suggests:

Demand growth

is the primary driver.

Now consider:

Backlog = +20%

Volume = stable

Headcount = stable

AHT = +25%

The likely area of investigation is different.

Now consider:

Backlog = +20%

Volume = stable

AHT = stable

Total capacity appears sufficient

but one specialist queue = +60%

Now the problem may be skill allocation.

A useful dashboard should make these distinctions visible.

Four Types of Backlog Problems

A practical diagnostic model is to classify backlog problems into four broad categories.

1. Demand Problem

More work is arriving than forecast.

Potential responses:

Capacity increase

Demand management

Temporary staffing

Revised forecast

2. Capacity Problem

Not enough effective productive time is available.

Potential responses:

Hiring

Scheduling changes

Overtime

Leave planning

Attrition replacement

3. Allocation Problem

Capacity exists but is not reaching the right work.

Potential responses:

Routing

Skill configuration

Queue redesign

Cross-training

Controlled assignment

4. Process Problem

The work requires more effort than necessary.

Potential responses:

Process redesign

AHT reduction

Automation

Quality improvement

Reduction in rework

Removal of unnecessary steps

Many real-world operations have a combination of all four.

A Practical Backlog Diagnostic

Suppose backlog increased from:

10,000 to 15,000

Management initially requests:

10 additional FTE

Before approving the request, examine the operation.

Demand

Forecast volume = 50,000

Actual volume = 52,000

Difference:

+4%

Some additional demand exists.

Productivity

Expected CPH = 5

Actual CPH = 4.9

Almost stable.

Workforce

Expected productive FTE = 70

Actual = 67

A modest capacity shortage exists.

Work Mix

Complex work increased from:

15% to 28%

This significantly increased handling requirements.

Quality

Rework increased from:

4% to 9%

Additional internal workload is being created.

Skills

Complex queue needs:

20 FTE

Only:

13 skilled FTE are available.

Now the picture is much clearer.

This is not simply:

“We need 10 more employees.”

The operation has:

• Slightly higher demand
• Slightly lower workforce capacity
• Much higher complexity
• Increased rework
• A significant specialist skill gap

The best response may combine:

targeted specialist hiring

cross-training

quality improvement

and:

rebalancing work allocation

rather than adding ten generic employees.

Use a Decision Tree Before Hiring

A practical management sequence is:

Is incoming workload above planned capacity?

If yes:

Determine whether the increase is temporary or structural.

If no:

Has effective productive capacity fallen?

If yes:

Investigate attendance, attrition, training, ramp and shrinkage.

If no:

Has AHT or work complexity increased?

If yes:

Investigate work mix and process changes.

If no:

Is backlog concentrated in a specific skill or queue?

If yes:

Investigate skill capacity and routing.

If no:

Has rework or repeated handling increased?

If yes:

Investigate quality and process design.

If no:

Is available capacity being allocated efficiently?

Investigate workflow, ownership and employee selection.

This sequence helps narrow the problem before increasing staffing.

When Hiring Is the Right Answer

Process improvement should not become an excuse to avoid necessary staffing.

Hiring is likely to be justified when:

Demand growth is real and sustainable.

Productivity assumptions are credible.

Process efficiency is reasonable.

Work allocation is functioning appropriately.

Quality and rework are under control.

Required capacity still exceeds effective available capacity.

In that situation:

Required Productive FTE > Sustainable Effective Productive FTE

and the difference is a genuine staffing gap.

For a detailed capacity calculation, read How to Calculate Workforce Capacity: From Headcount and Productive Hours to Required FTE.

When Process Improvement May Be the Better First Move

Process improvement deserves priority when backlog is primarily driven by:

• Excessive AHT
• Duplicate work
• Rework
• Manual handoffs
• Poor work allocation
• Unnecessary steps
• Repeated follow-ups
• Skill imbalance
• Avoidable administrative activity
• System inefficiency

The objective is not to avoid hiring.

It is to avoid hiring people to compensate permanently for a correctable process problem.

Sometimes You Need Both

Operational problems are rarely perfectly clean.

Suppose the operation is:

6 productive FTE short

and process improvement is expected to recover:

3 FTE equivalent capacity

Management may still need:

3 additional productive FTE

The correct answer is therefore:

Process Improvement + Targeted Capacity Increase

rather than choosing one or the other.

This is why the decision should be based on quantified drivers.

Translate Every Problem into Capacity Impact

One way to improve management decisions is to express operational issues in a common unit.

For example:

Higher absence = -3 FTE

Rework = -4 FTE equivalent

AHT increase = -6 FTE equivalent

Process improvement opportunity = +5 FTE equivalent

Skill shortage = -7 FTE in Complex Processing

Now management can compare different issues using a common capacity lens.

This does not make every estimate perfectly precise.

But it creates a much stronger discussion than relying only on anecdotal statements such as:

“The team feels overloaded.”

Build a Backlog Root-Cause Dashboard

A useful dashboard might connect:

The purpose is to help management connect the backlog outcome to its likely drivers.

Backlog Burn-Down Still Matters

Once the cause has been addressed, management needs a recovery plan.

A useful formula is:

Net Burn-Down = Completed Volume − Incoming Volume

Suppose:

Incoming = 2,000 per day

Completed = 2,400

Net burn-down:

400 cases per day

Current backlog:

8,000

Approximate clearance period:

8,000 ÷ 400 = 20 working days

This assumes demand and capacity remain reasonably stable.

If management wants to clear backlog faster, it needs additional temporary capacity or further process improvement.

Monitor Whether the Intervention Worked

Suppose management introduces:

• Cross-training
• Revised routing
• Process simplification
• Additional staffing

Do not stop at implementation.

Measure:

Backlog before

vs:

Backlog after

AHT before

vs:

AHT after

Skill gap before

vs:

Skill gap after

Rework before

vs:

Rework after

SLA before

vs:

SLA after

This creates a closed management loop:

Diagnose

↓

Intervene

↓

Measure

↓

Adjust

From “We Need More People” to Evidence-Based Capacity Decisions

A basic staffing discussion sounds like:

“Backlog is increasing. We need 10 more people.”

A stronger discussion sounds like:

“Backlog increased by 5,000 cases. Demand explains 1,500. Increased AHT explains another 1,200 of monthly capacity loss. Rework is consuming approximately 3 FTE, and Complex Processing has a six-FTE skill gap. We need targeted capacity plus process intervention.”

That is a much stronger management conversation.

The evolution is:

Backlog Increased

↓

Measure Demand

↓

Measure Effective Capacity

↓

Analyze Productivity and AHT

↓

Analyze Skills and Routing

↓

Analyze Rework and Process

↓

Quantify the Gap

↓

Choose the Intervention

↓

Measure the Result

Backlog should trigger analysis.

Not an automatic hiring request.

How Praevexa Can Support Better Operational Visibility

Praevexa focuses on practical tools and MIS approaches that help operations teams understand work, performance, quality and workforce availability.

The answer is not always “hire more.” Sometimes a relatively small improvement in AHT, automation or productive capacity can remove the staffing gap entirely. Praevexa Workforce Planner lets you test those scenarios before making the decision.

Compare staffing and productivity scenarios →

https://www.praevexa.com/WorkforcePlanner.aspx

Praevexa CaseFlow supports structured case-based operations, including work allocation, ownership, backlog, aging, SLA and operational reporting.

Praevexa QualityFlow provides structured QA information that can help organizations identify defects, quality trends and potential rework drivers.

Praevexa HRMS supports employee, roster, attendance and leave management, helping organizations maintain workforce-availability information.

These are separate applications, but the management questions they address are interconnected.

When backlog increases, leaders need to understand:

Is the problem demand?

Capacity?

Skills?

Productivity?

Quality?

Routing?

or:

Process design?

The correct answer may save the organization from either under-hiring—or hiring into an inefficient process.

Related Reading

How to Manage Backlog, Aging and SLA in Case-Based Operations

How to Calculate Workforce Capacity: From Headcount and Productive Hours to Required FTE

Why Your Operations Team Can Be Understaffed Even When Headcount Looks Sufficient

How to Forecast Headcount Requirements: Volume Growth, Hiring, Training, Ramp-Up and Attrition

FIFO vs Priority vs Skill-Based Work Allocation: Which Is Better?

Pareto Analysis in Quality Management