Why You Should Stop Staffing Your Call Center by Revenue (and What to Do Instead)

You’ve probably heard the old benchmark: “One CSR for every $1.5 million in revenue.”

It’s thrown around like gospel in the trades—plumbing, HVAC, electrical—and on paper, it sounds clean. Simple. Easy to plan for.

But let’s run that through the real-world blender:

If your company’s doing $120 million in annual revenue, that would mean you need 80 CSRs.

Eighty.

Eighty full-time, inbound-only customer service reps… just to “match” revenue?

Come on.

That’s not scalable. That’s not lean. That’s not even realistic. It's just plain stupid.

The Myth of Revenue-Based Staffing

That $1.5M/CSR benchmark works okay for companies doing under $10M—especially if they’re light on tech and the CSRs are booking everything manually.

But once you scale?

You need to think about:

• Actual call volume
• Call type mix (booking vs support vs fluff)
• Automation tools
• Channel options (text, chat, self-scheduling)
• Shift coverage and downtime
• Inbound vs outbound specialization

Trying to use revenue alone to determine staffing is like picking your wrench size based on the size of the truck. It might look close, but it doesn’t fit the job.

So, How Should a Call Center Be Staffed?

You build your team the way a contractor builds a house: start with the blueprint, not the budget.

Let’s break it down.

1. Start with Actual Call Volume

Don’t guess. Pull the reports. If you don’t have reports, fix that first.

Let’s say you handle 35,000 inbound calls per month across your branches. That’s about 1,750 per day.

Now ask: how many calls can one CSR handle well in a day?

• Booking-heavy? 60–80/day is reasonable.
• Scripted, transactional? You might get up to 100+.
• Mixed with admin clutter? Expect 50 or less.

Use that number to reverse-engineer how many CSRs you really need—not based on dollars, but on capacity.

2. Split Your Roles

Stop throwing everything at your CSRs. A high-performance team is specialized.

• Inbound CSRs – Focused on booking real leads
• SDRs / Outbound Reps – Rehash, tune-ups, membership campaigns
• Dispatchers – Routing jobs and managing tech productivity
• Admin Support – Permits, warranty calls, billing
• Inside Sales – Following up on unclosed estimates

Once you get big enough, Team Leads, QA, and Reclassifications

Each role has different metrics and expectations. One-size-fits-none.

3. Account for Shift Coverage

If you’re open 12 hours a day, 7 days a week, you’re looking at at least two shifts, and rotating weekend coverage.

You’ll also need to bake in:

• Breaks
• Sick days
• PTO
• Meetings and training

Pro tip: Most teams need 1.4 to 1.6x the number of seats to maintain full coverage.

Enter: The Erlang-C Calculator

Now let’s talk real math.

Erlang-C is a proven staffing model used in telecom, hospitals, and high-volume call centers. It answers the question:

“Given X calls per hour, Y average handle time, and Z desired wait time… how many agents do I need?”

✅ Pros of Erlang-C:

• Data-driven: It uses actual metrics, not guesses
• Predictive: Helps model different scenarios (e.g. peak vs off-peak)
• Efficient: Avoids over- or under-staffing
• Customizable: You can adjust for SLA targets (e.g. “answer 80% of calls within 30 seconds”)

❌ Cons of Erlang-C:

• Assumes constant call flow: Real life isn’t a perfect curve
• Doesn’t account for human breaks: It assumes agents are always available
• Ignores call complexity: Doesn’t separate sales calls from support calls
• Requires clean data: If your AHT (Average Handle Time) is junk, the result is junk

It’s a tool—not a silver bullet. But it’s way better than guessing.

Example: Erlang in Action

Let’s say your team receives 60 calls/hour, and your average handle time is 4 minutes. You want 80% of calls answered in under 30 seconds.

Run that through an Erlang-C calculator:

You’d need about 13–14 agents active during that hour to meet your SLA.

Add breaks, training, and coverage buffer = 16–18 total staff scheduled.

That’s much more strategic than “Well, we just hit $3 million, so hire two more CSRs.”

Take This Mental Model Instead:

• Staff for function, not revenue.
• Design roles around outcomes, not headcount.
• Use Erlang to model, not dictate.
• And automate everything you can that doesn’t generate revenue.

Final Thoughts: Build a Call Center That Drives Profit—Not Just Answers Phones

Your call center shouldn’t be a cost center. It should be a revenue engine.

You can:

• Reduce burnout
• Increase conversions
• Lower hold times
• Upsell memberships
• Protect your brand

And still have room to breathe

But only if you stop throwing bodies at the problem—and start building a system.

Revenue is the scoreboard. But calls, roles, and customer experience are the playbook.

You’ve got the people. You’ve got the tech. Now build the machine.

Here's an old video I made about this topic. 

I should redo it lol