It is worth it when customer messages arrive at hours when nobody on the team can answer in time. It is not worth it when what is missing is a sales process, because the agent handles the conversation and qualifies, but it does not create that process.
That is the short answer. The rest of this post shows the numbers I use to get there and the case where I advise against the investment.
The first sign: messages that arrive while the business is closed
In my experience, before the agent, the person who answered WhatsApp was the owner or one employee. That person has working hours. The customer does not: they write when they remember.
I measured this at two clients. At one of them, 38.9% of customer messages arrived outside business hours: 8,816 of 22,650, between January 19 and August 18, 2026. At the other, it was 25.1%: 20,845 of 83,151, between October 11, 2025 and October 16, 2026. The business hours used in the count are the hours each client registered, and I was the one who applied that rule to the data.
- First client38.9%
- Second client25.1%
In the lowest case, one in every four messages arrived while the business was outside its working hours. At the first of these clients, the agent’s median response time stayed between 3 and 6 seconds from April to July 2026, inside and outside business hours. I have no measurement from before the agent to compare that time with.
How much time this represents
At three of the businesses I serve today, the agents sent 7,661 replies between October 11 and November 10, 2026. Of those, 2,703 went out outside business hours, 35.3% of the total.
- 7,661replies sent by the agents
- 2,703replies outside business hours
- 35.3%of the total went out after hours
To turn this into hours, I use an estimate: about 45 hours for the replies outside business hours alone, on an estimate of 1 minute per reply. Adding up all the replies in the period, the same math gives about 128 hours. The premise of 1 minute per reply has no measured basis. It gives an order of magnitude, not a claim about how much time someone stopped spending.
Even with that caveat, the math helps you decide. If your WhatsApp gets few messages outside business hours, the agent solves little. If it gets many, that is where it works while the team rests.
When I do not recommend it
In a portfolio of smaller clients where I built customer service agents, the central lesson was this: there were clients without a sales process expecting the AI to be the salesperson and bring in profit. There was no defined sales process before the agent, and the agent does not create one. It handles the conversation and qualifies. The sale still depends on the business’s process.
Worth it when
- messages arrive that nobody answers in time
- the agent handles and qualifies the conversation
- someone takes over the customer afterwards
Not worth it when
- there is no sales process
- the AI is expected to be the salesperson
- nobody follows up on the handed-off conversations
That is why, today, I ask about the sales process before proposing an agent. A few questions help you run this test on your own:
- Who takes over the customer after the agent qualifies the conversation?
- Is there a clear next step, such as a quote or a booking?
- Does anyone follow up on the conversations the agent hands off?
If the answer is “nobody” or “I don’t know”, I would put the first investment into the process, not the agent.
What these numbers do not show
The numbers above measure service: a message that arrived and a reply that went out. They do not measure sales or conversion, because that data does not exist in the databases. There is also no period before the agent to compare with, since it was running from the first month of collection.
So I do not use these numbers to promise sales results. I use them to answer a simpler question: is there service work happening at an hour when nobody on the team is available?
How to tell if it is worth it for your business
Pick an ordinary week. Open the company WhatsApp and count how many conversations started outside your service hours. Then count the total conversations for the week and see what share fell outside working hours.
If that share is small, your problem may be somewhere else. If it is similar to the ones I showed here, an agent covers those hours without you hiring more people. Before you decide, also answer the questions about the sales process.
If you want to see how the WhatsApp agent I build works, the details are in services.