Supercharging Product Managers with Scopebrite 🤖- Building in Public
Welcome to the first #BuildingInPublic installment of Scopebrite, where we share the what, the why and the how we are building Scopebrite. Today we will be discussing the feature “Split Personas’’, which was identified early on and confirmed in one of our user interviews.
For some context, Scopebrite’s goal is to supercharge product and project managers through the power of Generative AI. Right now, given a description of what you are trying to build, Scopebrite successfully generates and suggests personas and other business stakeholders that are relevant to you, as well as user stories for those same personas.
In our case, due to the fact that the core of Scopebrite are Large Language Models (LLMs), the art of prompting, understanding and giving more context is key. However, defining the perfect prompt is an unreasonable goal. It is unfeasible to direct the LLMs to give the *perfect answer*, for all user inputs and for all user necessities. This is the challenge, which gives rise to opportunities. This post is about one of those opportunities.
If you are not familiar with Scopebrite and want more context, please go to scopebrite.com where we define its key features.
The Challenge
As part of any product, it is relevant to understand the personas, talk to them and define them. Scopebrite helps you out by generating personas given your product description. For example, as soon as you’ve given your description, you’ll end up with something like this:
It offers an overview of the suggested personas. For the sake of this example, this product is based on a social network focused on connecting people with experiences and events that they might care about.
Naturally you can add more, and explore further. In fact, when you drill down, you’ll end up with something like this:
It has a face, it has goals and challenges, we can easily identify with this persona. However, depending on the use-case, the label “Organizer” can be quite broad and it might not fit the needs of our users! That cannot be!
The solution?
It must be split!
The Solution
If it must be split, it shall be split! AI is here to help us.
In our early trials, we quickly validated that it was possible to leverage the same LLMs that we use to split a persona in two or more. As soon as we confirmed its feasibility, we had to figure out the experience that we wanted to provide to our users.
Option 1
Add a nice looking “Split” call to action in the top corner of each persona card. It offers easy access and it’s quick to find.
When the user clicks on it, a still-work-in-progress modal will pop up.
The user is free to decide which persona to keep (if any).
However, here are a few possible issues with this option:
- The accessibility of the ‘Split’ button might cause more cognitive load than actually being good. At the end of the day, it is another button that the user needs to decide if they want to click on it every time they go to the personas overview page.
- The modal is limited in functionality, if we want to allow the user to preview further the personas, do we use another modal? This might lead to modal madness.
Option 2:
Lets not have the Split in the persona card due to the above identified issues. Instead, let’s put the button in the details page of each persona, as shown below:
When the user clicks on it, a preview section will pop up.
The user can then accept or preview each persona and at the end confirm the choices or cancel everything. Since we have no modals in this version we can use them when the user clicks on ‘preview’ to drill down on the generated personas.
For example:
The Release
While things are still in the open, we are going to release this feature to our private beta users within the next few days. The team is more inclined to option 2, but we are curious about your thoughts and feedback! Maybe there is an option 3 that is more convenient ? Let us know! If you’re not in our private beta, we just opened a couple more seats, sign in at https://scopebrite.com/join-beta and we will reach out as soon as you can get it!
