Genus AI
Genus AI helps you automate creative production, target ideal audiences, analyse and improve ad performance, saving time and budget while boosting campaign results leveraging artificial inteligence.
The Brief.
I joined the team at Genus AI as a contractor to work with the CTO and CPO on an ongoing basis, assisting with feature improvements based on their findings. I was later joined by a senior product manager and a product designer, forming the core team that worked closely with the engineers and data scientists.
Initially, as the platform was not open to the public, feedback primarily came from the customer success and performance marketing teams. Later, we began to gather both qualitative and quantitative feedback through user testing.
My role in the project ranged from working on new and existing features to preparing high-fidelity mockups and prototypes, which assisted the team during grooming sessions.
Getting started.
As we opened the platform to the public, I helped map out the sign-in and registration flows, which served as a reference for the development team. One of the challenges we had was in regards to adding user accounts into existing teams. We used auth0 as the technology to hand that part but nevertheless assisted the existing logic with the appropriate flow.
We focused on obtaining a level of information that allowed us to set up the client account accordingly and promote the very important Facebook connection from the outset, this allowed to not land on an empty state dashboard screen but instead reveal paired data with the platform.
The Dashboard.
Upon landing on the platform, the dashboard highlights the three most important pillars for making the most of Genus AI. These include successfully setting up the Facebook connection, providing data sources to be used as audiences, and uploading content that can be used to build targeted ads.
Performance metrics.
After many iterations concerning the level of information displayed on the dashboard, this section became paramount to customer success managers. It served as a reference in sales calls and allowed for an at-a-glance analysis of the platform’s effectiveness.
By offering the capability to create a myriad of filters tailored to the needs of various clients, and providing the ability to quickly compare the performance of campaigns run independently versus those run through Genus AI, customers can swiftly make strategic marketing decisions.
As the definition of success in campaigns can be interpreted in different ways, we came up with the option to configure one of the columns with a custom performance evaluation indicator, allowing to define the primary metrics and their values.
We provided an overview of top performing ads in the dashboard - This fixed the need to go into each campaign, ad set and ad to reveal insights. We opted for designing the ad preview inside a modal that could be triggered from anywhere in the platform and contained all the associated data.
Content.
One of the core features of Genus AI is its ability to gather creative insights. Upon uploading an image or video to the platform, these are automatically scored according to archetypes and given a predictive score on ad appeal and click-through rate.
These creatives form the basis for the creation of ads in campaigns, and the scoring mechanism supports suggestions based on the intended targets.
By providing creative insights at folder level but also at library level, we give the ability to make strategic decisions in regards to the production of future content and also in regards to organising folders based on a given criteria.
The library section with the scoring functionality provided a feel for what type of audience content performs better. As this feature got a substancial amount of traction and due to the high volume of creatives to handle, we decided to provide advanced filters - This allowed to shortlist creatives to later be used in ads and streamline the ad set creation.
Enrichment report.
The platform has the capacity to model a target audience using a no code feature and, for US based audiences, enrich datasets.
As there was a strong demand to visualize the output of these audiences, we designed dashboards that showcase the breakdown of the dataset in regards to personality - By communication archetypes and personality traits.
Part of the report are also demographics - This allows to see at a glance the distribution of the audience in terms of gender, age, location, income, education and interests.
This feature was used extensively when demo-ing the platform to new leads as it provided the capability to compare your audience with the US national average.
Generate the best video ads for a given audience.
Beyond managing existing assets, the need to provide asset creation capabilities surged through client feedback. The engineering and data science teams started working on a proof of concept to generate videos based on inputing a number of creatives (image or video), scoring them using the proprietary Genus AI engine and output a video ad that’s optimised towards a goal.
As this became the priority feature to launch before the platform opened to public, I mapped out the video creation journey and worked on low-fidelity wireframes to support the joint effort to align the technology with the final interface.
We decided to split the video creation flow into two moments - First, the template selection that serves as a starting point to define the video style, transitions and animations. This together with defining the end goal for this video.
Second, you jump into the video builder interface and start tweaking details. We placed the toolbar on the left, in line with patterns of creative editing software, for familiarity reasons.
Final note.
The work I’ve done together with the team at Genus AI enables brands worldwide to generate product catalog images, seed audiences, copy and video to help grow efficiently across multiple social channels. The platform reached 10 million product images generated in 2023 alone. Genus AI has successfully raised $6m USD so far.