How I Use ChatGPT to Uncover Jobs to be Done
It has been my understanding that ChatGPT was trained on text content from the internet. Which included social media posts and blog posts. People typically discuss their problems/experiences via social media and blog posts. Therefore, I theorized that ChatGPT may be utilized as a tool to study the jobs customers are trying to get done when they purchase a product. In other words, study the problems customers are trying to solve when purchasing a solution.
What I first did was create a prompt that explained the details of “jobs to be done,” theory. I then added to the prompt an explanation of what market and customer is being studied. This way ChatGPT would understand the perspective I want in its output.
For example, I did a study on the transportation market by focusing on commuters who are trying to commute to work.
At the end of my JTBD theory prompt, I added the following prompt;
[Understand the following: (A commuter is a person who travels some distance to work on a regular basis.)]
[Market definition: ({commuters} + {commute},Contextual clarifier: {United States transportation market})]
I then combined the two prompts. Following the combined prompts, I ended my input by asking it to identify ten emotional jobs customers are trying to get done when commuting. So the end portion of my input would look like this;
Understand the following: [A commuter is a person who travels some distance to work on a regular basis.] Market definition: [(commuters) + (commute),Contextual clarifier: (United States transportation market)] List ten emotional jobs commuters are trying to get done when commuting.
The output received is as follows;
This gives us an idea of what may be the emotional jobs commuters are trying to get done when commuting. However, it is important to note that these “jobs,” must be verified with job performers before selecting one to shape your product design around.
For one client, I tested the validity of my findings using content. I shaped content messaging around my findings and determined which message generated the greatest results. My logic was that if the messaging resonated well it was most likely a job our customer was trying to get done when using our products. Shaping content messaging around the uncovered and validated anticipated outcomes is how I was able to cultivate their YouTube community presence, resulting in a Y2Y +90% subscribers, +362% views, +84% watch time, +69% impressions. Resulting in 91.3% yearly website engagement rate from YouTube traffic.
What are your thoughts on my experiments? Have you tried conducting this kind of research yourself? Do you want to know more about my jobs to be done theory prompt?Reach out to me and let me know?
