"If Message Does Not Contain" keyword filter is not working

Hi, ​@dandaya!

As far as I know, this trigger condition will trigger EVERYTHING except messages that doesn't contain the keywords you set to it. 

Best practice is using very specific words, made up codes and other unique keywords to your automations.

Documenting it is also super advised! Once you have some years of automations, it can become chaotic if you “let it loose”. 😅

Maybe it's time to review your entire automation list and organize it.

And if you're using more than one keyword to the same automation, you can always add a new trigger to the same automation and keep it tidy.

Even if only a part of your automation is similar, you can use the “Start Automation” step and send everything to the same part. This is wonderful when it comes to automation maintenance. 🤩

Hope this helps! 

🙌

Hey, ​@dandaya 

When you want to use “message doesn’t contain” trigger, you have to set up another trigger. Something like this:

First, choose one of these triggers (message contains/message contains whole word/message begin with)

Then, choose the trigger “message doesn’t contain”

Let me know if this helps 🙂

Hi ​@Fabio Gaulke 

Thank you for your reply.  Yes that filter is exactly what I need.  For context, I use ManyChat as a bridge between my custom AI Agent and our Human Agent. 

I have set up ManyChat so I can direct all initiated chats to my AI Agent that recognizes any question about my company (with any keyword for that matter) and answers accordingly.  My intention was to create an automation having no keywords that simply directs all conversations to that AI Agent.

Then I placed an automation with a keyword that “ends” the AI Agent conversation and passes it back to a human that uses other automations within ManyChat.  It’s a very clean and supposedly simple set up.  

Was the “If the message does not contain” keyword filter intentionally disabled because of the best practice model  that you described above?  If so, it would probably be best not to  include it in the choices. It was quite frustrating.

I tested my account with a specific keyword to see if it works  and eventually I activated my paid account with the thought that I had the option to use the “If Message does not contain” filter only to find out it did not work.  Since then, I’ve had to put a trigger that contains basically 100+ keywords that targets every iteration of every question that can be asked about my company.  Its working well but it made my set up process very long and tedious.

In any case, I would still suggest to enable or fix this filter because AI Agents such as mine are inevitable and for the most part, we will not need to use keywords in the future as these AI Agents have to ability to understand and answer questions and human assistance will only bee needed for higher value conversations after initial contact.

Let me know if that filter is permanently disabled so I know what to expect. 

Again, thank you very much for the reply - appreciate it.

Hey, ​@dandaya 

When you want to use “message doesn’t contain” trigger, you have to set up another trigger. Something like this:

 

First, choose one of these triggers (message contains/message contains whole word/message begin with)

Then, choose the trigger “message doesn’t contain”

 

Let me know if this helps 🙂

​@rodrigo_silvano 

Thank you very much for this.  Now I understand that it is sort of a sub-filter.  I was hoping to use it as a way to trigger any keyword to direct any conversation to my AI Agent.  Please see my answer above to ​@Fabio Gaulke.  It would be great if we can use it as a main filter for the reasons and use case I outlined above.

Again thank you!

Dustin

Hi everybody! Thanks for your replies, it helped! 
Question: I’d like to have a flow when there are three conditions:

1. Message from my Ad set #1

2. Message from my Ad set #2

3. Everything else.

Is there a way to put all of these three condonations into a single flow? I didn’t see how I can do it.
For now, I’ve created one set for the Message from my Ad set #1 and another set for the Message from my Ad set #2. Now I have to build an automation for the rest of the cases and I don’t want to use the “message exclude” rule in order NOT to miss anything. Any advice here please? Very much appreciated! 

Hey, ​@NGoro

Your message that is sent from ad set #1 is the same message that is sent from ad #2? Because if it’s not, why would you build a big automation instead of two small ones?

Now, I’m assuming that the ads triggers sent two different messages:

You’ll need to build 3 small automations and the main one.

1. First automation:
   • Ad set #1 trigger
   • Set the first message that the user will see
   • Add a tag to the user. For example: “ADSET1”

2. Second automation
   • Ad set #2 trigger
   • Set the first message that the user will see
   • Add a tag to the user. For example: “ADSET2”
3. Third automation
   • Set your default reply
   • Add a tag to the user. For example: “DEFAULTREPLY”

4. Fourth automation
   • Set your triggers based on tags
   • Then, add a condition block to separate the flows based on the user’s condition

Doing that, you can concentrate your different triggers in one automation.

Let me know if this helps 🙂

Thanks rodrigo_silvano for your response! 

Your assumption is correct, there are two different Ad sets.

1. Ad set #1 - sending a file #1 to user → add a tag & close the conversation. 
2. Ad set #2 - sending a file #2 to user → add a tag & close the conversation. 
3. I would like to have a default reply that will be triggered by anything except of Ad set #1 & Ad set #2.
   In my default response I  would like to EXCLUDE users from Ad set #1 & Ad set #2. Meaning: if the user came in the way other than  Ad set #1 & Ad set #2 → get the default response. I cannot set a trigger based on a tag, since in the beginning of the session there is no tag yet.
   What would be the simplest way to send a user a default response on the following condition: the user didn’t come from Ad set #1 nor the Ad set #2?​​​​​​​Thanks & very much appreciated! 

Thanks rodrigo_silvano!

My original response hasn’t been sent for some reason. Your response is very helpful. 

Thanks again,

Nadia