Formula for Sizing Your WhatsApp Support Team: Messages per Hour, Agents Needed, and Calculation Table

What you will solve with this page

The business WhatsApp doesn't answer itself. Sooner or later the time comes when messages pile up, customers complain about the delay, and someone says "we need more staff". But how many staff? Hiring too many is throwing money away; hiring too few is staying the same. This page gives you the formula to calculate it with numbers from your own business, not with wild guesses.

The formula is simple: you need to know how many messages you receive per hour, how long an agent takes to resolve each one, and how many hours per shift each person works. With that, the calculation comes out in two minutes. Then you can adjust it for peaks, breaks, and tasks that aren't answering WhatsApp.

The three data points you need

Before doing any math, you have to gather three numbers. If you don't have them, don't make them up: measure them for a week and then apply the formula.

• Messages per hour (M): how many messages come in on average in one hour. It's calculated by dividing the total messages of the week by the hours you attend. If your business has hours from 9 to 18, that's 9 hours per day, 45 per week.
• Response time per message (T): how many minutes it takes an agent to resolve a message from start to finish. It includes reading, thinking about the answer, writing, and waiting for the customer's reply if it's a conversation.
• Productive hours per agent (H): how many hours per day an agent is actually available to attend. Nobody performs for 8 straight hours: you have to deduct breaks, meetings, and administrative tasks. A realistic number is between 5 and 6 productive hours per 8-hour shift.

With those three numbers, the formula is as follows.

The formula and how to read it

The number of agents needed (A) is calculated as follows:

A = (M × T) / (H × 60)

Where M is messages per hour, T is minutes per message, H is productive hours per agent, and 60 converts hours to minutes. The result tells you how many agents you need at peak time. If the number is, say, 2.3, round up: you need 3 agents to cover that moment with margin.

Example: you receive 40 messages per hour, each agent takes 4 minutes per message, and each has 5 productive hours per shift. The calculation is (40 × 4) / (5 × 60) = 160 / 300 = 0.53. In other words, one agent is enough, but just barely: that agent will be busy 53% of their time. If the number goes above 0.8, it's no longer healthy: the agent has no margin for unexpected events and response time will skyrocket.

Quick calculation table

So you don't have to do the math every time, here's a table with pre-calculated results for typical combinations. Find your case or interpolate between rows.

The table assumes 5 productive hours per agent. If in your business it's 4 or 6, adjust: fewer productive hours, more agents; more hours, fewer agents.

How to measure your business data

If you don't have the numbers, don't guess them. Spend a week measuring with tools you already have. Most customer service systems (including Wando, which shows you message statistics) give you the total messages per day and per hour. With that, you can calculate the average per hour.

For response time, the easiest way is to time a sample: note how long each agent takes to resolve 20 messages and calculate the average. If the messages are simple (inquiries about hours, prices), they will take 2 or 3 minutes; if they are complaints or complex inquiries, they can take 8 or 10.

Adjustments for peaks and critical hours

The formula gives you the average, but WhatsApp doesn't behave like an average: there are peak hours in the morning, in the afternoon, or after a social media post. If the peak is double the average, you need double the agents at that time, even if the rest of the day they are at 30%.

Two ways to handle peaks: assign agents in staggered shifts (not everyone starting at the same time) or use auto-responses to say "we'll get back to you in a moment" so the customer knows their message was received. Neither replaces an agent, but they help the wait not feel like abandonment.

Common sizing errors

• Using the daily average instead of the peak hour: if you get 60 messages in one hour and 5 in another, the average is 20, but in that peak hour you will need 3 agents, not 1.
• Not discounting non-productive hours: if you count 8 hours per agent and they actually deliver 5, the calculation falls short and the team gets overwhelmed.
• Underestimating time per message: many only count the time to write the response, but you have to add reading, waiting for the customer's reply, and messages left mid-conversation.
• Rounding down: if the result is 1.8 and you put 1 agent, that agent will be overloaded and the response time will skyrocket.

How to know if the sizing is right

After applying the formula, check with two indicators: the average response time and the percentage of messages left unanswered during the day. If the response time is more than 5 minutes during normal hours, you probably need more staff or the time per message is higher than you thought. If unanswered messages are more than 5%, the problem is structural: improving the wording is not enough, you need to add hours or agents.

Also look at each agent's workload: if one is at 90% of their time and another at 20%, the problem is not the total number of staff but the distribution of tasks. In that case, the solution is to reassign, not to hire.

Complete applied example

A clothing store receives 120 messages per day from Monday to Saturday, during hours from 10:00 to 19:00 (9 hours). The average is 120/9 = 13.3 messages per hour. Each agent takes 4 minutes per message (size, price, stock inquiries). With 5 productive hours per agent, the formula gives (13.3 × 4) / (5 × 60) = 53.2 / 300 = 0.18. One agent is enough with plenty of margin.

But on Saturday afternoon, after an Instagram post, 60 messages arrive in one hour. During that peak, the formula gives (60 × 4) / (5 × 60) = 240 / 300 = 0.8. With only one agent, that Saturday will be at the limit. The solution is not to hire a second full-time agent: it is to have someone cover Saturday afternoon, or use auto-responses for simple messages and let the agent focus on those that require attention.

What happens when the result is less than 1

If the formula gives you 0.3 or 0.5, it doesn't mean you don't need anyone: it means that person won't be busy all day handling WhatsApp. The most common mistake is thinking that one agent is enough and that's it, but that agent also has to eat, go to the bathroom, and handle other tasks. The practical rule is: if the result is less than 0.7, one agent is enough, but you need to assign them other tasks for the dead hours. If it goes above 0.7, you're in a risk zone: any unforeseen event (a long message, an insistent customer) overwhelms them.

Multi-agent: how to distribute without stepping on each other

When the calculation gives you 2 or 3 agents, the problem stops being how many there are and becomes how they are organized. If everyone handles the same inbox, two might answer the same message, or no one takes charge of the ones left hanging. The practical solution is to divide by simple criteria: by schedule (one covers the morning, another the afternoon), by type of inquiry (one sales, another complaints), or by message status (one for new ones, another for those already in conversation). Tools like Wando allow multi-agent precisely so each one sees what they need to handle and they don't step on each other.

The uncomfortable question: what if the problem isn't the number of people?

Before hiring someone, check if the time per message (T) can be reduced. If your agents take 8 minutes per message because they write long responses by hand, a template system or AI suggestions can bring it down to 3 or 4. That's not a minor detail: if you go from 8 to 4 minutes, the formula tells you that you need half the agents for the same volume. Many times the problem isn't a lack of staff, but that the staff is wasting time on tasks that can be automated or streamlined.