Hi everyone!
Do you want to know a little secret about how you can use AI to do your dirty work? At first glance, this might sound like a nefarious scheme, but I promise it’s actually a genius way to protect your team’s valuable time and ensure high-quality work. Here’s how to let Asana AI play the “bad cop” in your workflow:
Imagine a content team at Phoenix Corp. They work hard, they produce world-class content, and they value time for heads-down work above all else. Everybody at Phoenix Corp loves them.
They receive content production requests into their intake project, where members of various teams at Phoenix Corp can request copy for social media posts, help articles, blog-style publications, and more.
Alan hasn’t provided all the information that the Content team needs. And therein lies the problem.
The quality of the work that they can produce is underpinned by the quality of the information provided to them by the requester. The content team can’t begin work unless they’ve got all the context they need, and when they don’t get it up front, they have to waste time and energy reaching out for more info, and following up on unclear expectations.
And they hate that.
The setup
Enter the “bad guy”, an AI-powered quality checkpoint automation which intercepts incoming requests before they ever reach the content team. All created with AI Studio.
- When a task moves to the QA Checkpoint section of the intake project, the Asana AI springs into action. All automatic. No human effort required at this point, so the content team members can keep diligently typing away.
- The AI checks if the request includes three vital pieces of information:
a. A detailed description of deliverables (maybe even including a minimum word count
)
b. A due date for the request
c. Information on the target audience
We’ll show exactly how to implement this checkpoint at the bottom of this post.
The result
-
If the request passes QA
, it automatically moves to the Content Creation stage. Only at this point are the content team informed about the request. (Of course, they can see anything they want within the project, but this will be their first notification). -
If it fails QA
, it gets diverted to a Needs More Info section, where the Asana AI (not a human - too busy) automatically asks for the missing details. When the requester replies, the task is routed to the QA Checkpoint again, and (hopefully) can move to Content Creation with all the relevant details provided.
Alan’s forgotten something in his request. But no problem, the AI will steer him in the right direction.
Why this is brilliant
The Asana AI handles the “mean” work of rejecting incomplete requests 
OK, the real benefit is obviously that human team members don’t need to spend time chasing information and being interrupted by low-quality submissions, but it’s nice to be nice.
In addition to protecting your team’s time, requesters actually get immediate feedback and quality standards are consistently enforced. And if nobody has to be the bad guy, well that’s just good fortune!
P.S. Want to implement this scheme yourself? Just set up the following rules with these triggers and actions in your Asana project. The Asana AI will happily play bad cop while you focus on the work that matters.
First, set up the QA check rule.
The trigger is when the task is moved to the QA Checkpoint section of the project. There are 2 conditions with corresponding actions.
- Check if the request has enough information
- Check if the request does not have enough information
If there is enough information, the actions performed are to move the task to the Content Creation section, and to assign the task to a member of the Content team (Michelle in this example).
If there is not enough information, move the task to the Needs More Info section.
Next, set up the Discovery rule.
- The trigger is when the task is moved to the Needs More Info section of the project.
- There is no condition in this rule.
- The action is to add a comment, which simply includes the Fill by AI variable.
Finally, set up the QA loop rule.
The trigger is when a comment is added.
- There are 2 conditions, check if the task is in the Needs More Info section, and check if the last comment was not made by Discovery. This will capture the requester’s response when they reply.
- The action is to move the task back to the QA Checkpoint section, to go through the QA check again.
Here is a question for you in relation to your demo.
