Live chat staffing calculator
Find how many agents you need for live chat, phone and email. Erlang C with chat concurrency, shrinkage and occupancy, plus what it costs and how much an AI agent takes off. Free, no sign-up.
Made by Anatolii Ulitovskyi 
Fact-checked by RH Fardin Updated Fewer chats in the queue
Let an AI agent answer the routine questions and keep your team for the conversations that need a person.
CRMsoftware.pro answers customers on your website, WhatsApp, Telegram and email in seconds, day and night, and hands off to your team with the full conversation when a person is needed.
- AI agent trained on your website, documents and policies
- Hands off to a teammate with the full context
- Every chat becomes a contact and a deal in your pipeline
Free plan, no credit card. Paid plans from $19/month, billed yearly.
Plan by hour
Enter your contacts for each hour you are open. The example pattern is a placeholder: replace it with your own numbers from your chat or phone reports.
The method
How the staffing calculation works
Live chat and phone queues use Erlang C, the standard queueing formula for contact centers. It finds the smallest number of agents that meets your service level.
Erlang C then works out the chance that a customer has to wait and how long, for each possible number of agents. The calculator picks the first number that answers your target share of contacts within your target time without pushing occupancy above your limit. Finally it adds shrinkage, because agents are not available every minute they are paid.
40 chats an hour
40 × 10 min ÷ 1.5 = 4.44 Erlangs
7 agents logged in answer 82.9% of chats within 30 seconds. With 30% shrinkage, schedule 10.
60 calls an hour
60 × 6 min = 6 Erlangs
9 agents logged in answer 83.4% of calls within 20 seconds. With 30% shrinkage, schedule 13.
300 tickets a day
300 × 8 min = 40 hours of work
At 85% occupancy and 30% shrinkage, an 8-hour team needs 9 people.
Chat is not phone
Chat concurrency changes the answer
An agent can handle more than one chat at a time, so chat needs fewer agents than the same number of calls.
The calculator divides chat handle time by concurrency before it runs Erlang C. Use the average number of chats your agents really handle at once, not the maximum your chat tool allows. Contact center planners usually see between 1 and 2. Going higher often makes each chat longer, so the saving is smaller than it looks.
With the example above, 1 chat at a time needs 10 agents logged in. 1.5 at a time needs 7, and 2 at a time needs 6.
Agents who juggle chats often type slower in each one. Use the slowdown field to make every extra chat longer, for example 15%. At 2 chats at once and 15% slowdown, each chat counts as 11.5 minutes instead of 10, and the example needs 7 agents logged in instead of 6.
Real availability
Shrinkage and occupancy
Two numbers turn the theoretical answer into a schedule you can actually run.
Paid time with no contacts
Breaks, training, meetings, sick days and holidays. Most staffing tools start at 30%. Use your own figure when you have it.
Busy time while logged in
Above about 85% for long periods, agents burn out and quality drops. The calculator adds agents until occupancy is at or below your limit.
The promise to customers
A common phone target is 80% of calls answered within 20 seconds. Set the target you actually want to keep for each channel.
When customers leave
Erlang A: staffing with abandonment
Erlang C assumes every customer waits as long as it takes. Real customers give up, so Erlang C can ask for more agents than you need.
Enter the average time a customer waits before leaving, and the calculator switches to Erlang A. It reports the share of customers who leave and counts them as not answered in time, so your service level stays honest. A shorter patience usually means fewer agents but more lost chats or calls.
Find patience in your chat or phone reports: the average waiting time of contacts that were abandoned is a good starting point.
No live queue
How to staff email and tickets
Email does not need an instant answer, so it is planned by daily workload instead of Erlang C.
This assumes tickets are answered on the same working day. If you promise a 48-hour reply, you can spread a peak day over two days and staff for the average instead.
What changes with AI
How an AI agent changes staffing
Every conversation an AI agent fully resolves never reaches the queue, so the workload falls.
Enter the share of conversations you expect an AI agent to resolve without a person. The result table shows agents and monthly cost at 0%, 20%, 40% and 60% so you can compare scenarios. The saving is not linear: queues need a minimum team, so the first conversations you remove save less than the next ones.
Measure the real share after launch and update the number. Conversations the AI hands over still count as human workload.
Spreadsheets
Staffing formulas for Excel and Google Sheets
A1 = contacts per hour, B1 = handle time in minutes, C1 = agents logged in, D1 = target seconds.
Workload in Erlangs
Chance a customer waits (Erlang C)
Service level
People to schedule with 30% shrinkage
FAQ
Staffing questions
How many live chat agents do I need?
Multiply chats per hour by the average chat length in hours, divide by how many chats an agent handles at once, then run Erlang C for your service level. At 40 chats an hour, 10-minute chats and 1.5 chats at a time, 7 agents logged in answer about 83% of chats within 30 seconds.
What is Erlang C?
Erlang C is a queueing formula from telephone engineering. It estimates the chance a customer waits and the average wait for a given workload and number of agents. Contact centers use it to plan staff for phone and chat queues.
How many chats can one agent handle at the same time?
Most teams average between 1 and 2 at once. Simple questions allow more; complex support allows fewer. Use your measured average rather than the maximum set in your chat tool.
What is a good service level for live chat?
There is no single standard. Many teams aim to answer most chats within 30 to 60 seconds, because the customer is waiting on the page. Pick a target you can keep at peak times.
What is shrinkage in a contact center?
Shrinkage is the share of paid time agents are not available for contacts: breaks, training, meetings, sick days and holidays. Most staffing tools use 30% as a starting point.
Why does the calculator use the peak hour?
Staffing for the busiest hour keeps service level when it matters most. For a full schedule, run the calculator for each hour of the day and use fewer agents when it is quiet.
Does Erlang C work for chat?
Yes, with an adjustment. Dividing handle time by concurrency gives a good estimate for planning. It assumes customers do not leave the queue, so it can slightly overstate the agents you need.
Can I download a live chat staffing calculator for Excel?
Yes. Fill in the hourly plan and click Download for Excel. The file opens in Excel or Google Sheets with your inputs, agents for each hour, total FTE, monthly cost and the AI scenarios. The Excel formulas section above shows how to rebuild Erlang C in a spreadsheet.
How do I calculate FTE for a contact center?
Add up the agents you need to schedule for every open hour of the day, multiply by the open days per week and divide by 40 hours. The hourly plan does this for you and includes shrinkage.
How does an AI agent reduce staffing?
Conversations an AI agent resolves never reach your team, so the workload falls. Enter your expected share in the calculator to see agents and monthly cost at 0%, 20%, 40% and 60%.
References
Sources
- Chat concurrency in forecasting: Call Centre Helper
- Customer service representative median pay, May 2025: U.S. Bureau of Labor Statistics
- Erlang C formula: Wikipedia, Erlang (unit)
More free sales tools
Answer every customer without growing the team.
CRMsoftware.pro answers your customers with AI, hands off to your team when needed and puts every lead in a pipeline. Free plan, no card.