⬅︎  Back home

Astral

AI-Assisted Platform Content Generation
Highlights
  • โœจ  An AI-assisted layer supercharging already existing material
  • ๐Ÿงช  I aligned the product to market fit through metrics-driven iteration
  • ๐Ÿ™‚  Focus on humanized interactions with contextually relevant content
  • ๐Ÿ› ๏ธ  Incorporated component system utilizing NextUI and React

A win-win

The value of Astral is being able to convert an existing content library and automate future publications (e.g. podcast episodes, blog posts) into additional engaging content - growing the audience while saving time and money.

First iteration

MVP: An Astral-hosted space with bite-sized content bits from the library

Initially, Astral's value came in the form of a dedicated feed. To enable deeper engagement with the content, a 1-to-1 chat experience was created, replicating the tone of voice of the Expert. This allowed Users to access in-depth knowledge at any time, without having to navigate through endless publications.

1. Chat

Chat interactions includes further value relating to the topic when possible

The responses are not only in the same tone of voice of the Expert, but also include additional value such as contextual upsells & furthers links to source material.

Upsell Upsell
Upsell
Adding the Expert's Astral to your homescreen means easy access at all times

However, this meant that the Expert had to segment their traffic to a separate space. To address this, the focus shifted to integrating Astral directly into the Expert's existing platform.

2. Launcher & Overlay

Widget Widget
We A/B/C tested open rates using several factors, including messaging, button designs, menu options, placement, and more

This embedded solution enables the audience to interact with content that's not necessarily present on the website, i.e. ingested from external feeds and publications.

Based on engagement metrics, various overlay variations were tested and iterated upon to determine the most engaging elements, such as Highlights and bite-sized Summaries, leading into a contextually crafted up-sell funnel or email capture.

๐ŸŽฏ

Right-center  variant clearly won with a higher probability of being opened. This also prevented potential collision with cookie banners on mobile.

3. Onboarding

An initial variant of the Onboarding flow, leading to the Self Service area. Shortly after simplified to three steps

To simplify the onboarding process and allow Astral's customers to set it up and start benefiting with lowered effort, numerous user journeys, flows, and prototypes were crafted to land on the most intuitive experience. A key challenge was to avoid overwhelming the User with the initial step of adding existing content to get started.

Onboarding

4. Self Service

Dashboard Dashboard
The overview of audience engagement on different topics also serves up content opportunities by weighting it to existing material

After onboarding, the focus shifted to understanding why Experts would visit the dashboard in the first place. Through research, it was determined that Insights and Engagement Metrics became the focal point, while the Library section was less prominent after the initial setup.

5. Content Remix

Remix to any amount of different types based on only 1+ piece of published content

Eventually, after several iterations to find the most useful market fit through demos and testing, it was discovered that generating Content Remixes was the most crucial user case that got people excited. This led to a pivot, highlighting Astral's ability to transform content, such as turning a podcast into a course or show notes into attractive social media posts.

workflow
Behind the scenes the Workflow Editor enables any custom setup of inputs to outputs

To support this new focus, repeatable UX and UI patterns were established, ready to have remixing workflows attached to them. I then iterated on finding the easiest way to manage content generations, automations, and library structure.

Astral