What happens if two customers write on WhatsApp at the same time?
Many customer service teams that use Odoo and Make face the same problem: they have managed to get their agent to reply on WhatsApp automatically, but they fear that two simultaneous conversations may cross or that the system will fail when handling several chats at once.
This article shows you how to properly configure your scenario in Make so it can serve several customers in parallel, without losing control of the conversation or mixing contexts.
The problem: simultaneous service without chaos
When we automate WhatsApp service from Make, each incoming message becomes a new execution of the scenario. If we do not properly separate the context of each customer, messages can get crossed, generate wrong responses or break the user experience.
Real story: Clara, an Odoo consultant at a services company, set up a bot that answered frequently asked questions on WhatsApp. Everything worked well until two customers wrote at the same time and the bot replied to one with the message meant for the other. The error caused distrust. The cause: she was not isolating the context of each conversation properly.
Solution 1: Ensure parallel execution in Make
Make can process scenarios in parallel, but this option must be explicitly enabled:
- Go to the scenario that manages incoming messages.
- Click the gear icon (⚙️) in the upper right corner.
- Find the option “Sequential processing” and make sure it is set to No.
- Save the changes.
Tip: you don't need to configure this in the webhook module, but at the level of the complete scenario.
Real story: Luis, an Odoo partner, had a bot that worked well... until two users wrote a few seconds apart. By changing this option, he got Make to process both messages in parallel without collisions.
Solution 2: Use unique identifiers per conversation
To keep conversations separate, each message must be associated with a unique thread_id. Generally, you can use a combination of the customer's WhatsApp number and a timestamp, or an ID generated by your system.
When you save data such as conversation history or state, use that thread_id as the key. This prevents two sessions from sharing the same “memory space”.
{
"thread_id": "+51987654321_202507181435",
"last_message_time": "2025-07-18T14:35:00Z",
"status": "open",
"notified_no_response": false
}
Real story: Ana, a team lead on a sales team, wanted to follow up on active conversations. Her bot used only the phone number as the key, which caused confusion when the customer wrote days later. With the thread_id implemented, each conversation was clearly delimited.
Solution 3: Automate closing conversations due to inactivity
So the bot does not carry old contexts forward, you can set up a scenario that checks open threads at regular intervals (for example, every 5 minutes) and applies these rules:
- If more than 5 minutes have passed without a reply from the customer and they have not yet been notified, send a message such as: “Are you still there?”
- If more than 15 minutes have passed without a reply, close the thread and delete its history.
Example flow in Make:
- Scheduler module: runs every 5 minutes.
- Search Data Store records where
status = "open". - Iterate over the results.
- Evaluate whether more than 5 or 15 minutes have passed since
last_message_time. - Send the follow-up message or close the thread, as appropriate.
// Decision pseudocode
if now - last_message_time > 15min:
status = "closed"
context = null
elif now - last_message_time > 5min and not notified_no_response:
send("Are you still there?")
notified_no_response = true
Real story: Martín, a process analyst, noticed that many leads did not reply after the first contact. After setting up this flow, the return of customers who resumed the conversation after the “Are you still there?” message increased by 28%.
What do I gain by applying this?
By implementing these improvements:
- You avoid cross-conversation errors.
- You provide smoother and more professional service.
- You automate closing old threads, which optimizes resources.
- You get closer to a "human" experience, without needing human agents.
Conclusion
Automatic WhatsApp service is a powerful tool for Odoo companies, but only if it is well structured. Separating context per conversation, allowing parallel executions and controlling inactivity are key steps to deliver a professional and scalable experience.
Implementing these improvements does not require being a programmer. Just know the Make tools well and organize the information correctly. Start with small steps and you will see how your customer service improves tangibly.