Marketeer
Description
Marketeer is a web app that allows small to medium sized businesses to create targeted ads towards multiple groups of consumers across multiple platforms.
We created this project as a two-man team with the help of an industry mentor. I designed and built wireframes, performed user testing and user research, and created prototypes initially in Figma then finally programmed a working web app prototype.
problem
Marketing is an expensive and time-consuming endeavor. Marketing across multiple platforms can be difficult as each platform has its own unique quirks and user bases. This is even more difficult when trying to simultaneously court different types of customers. Splitting up resources to market to multiple groups of user bases is a dilemma for large businesses. For small businesses the ability to do so is almost out of the question. We were specifically looking for areas of tedium where AI could provide value rather than overengineer an expensive solution.
Identified Issues
The solution we arrived at was to leverage AI to help automate the time-consuming aspects of market research, and content generation. By doing so, we allow our users to generate images and text aimed specifically towards each market segment for every desired advertising platform.
The user either manually enters details about their business and customers or provides a link to their website and to review sites of their website (google reviews and/or Yelp).
Marketeer scrapes information from these reviews and from the business website into a description of your business, then categorizes the customers into market segments. The details of all of these things can be manually changed as desired.
Marketeer then generates images and text based on the platform designated (google-AdSense, Instagram, Facebook, etc.), the market segment selected, relevant business details, and input from the user.
The Working Protoype I created
Methodology
We gave some time for ideation, competitive analysis, and research initially. Following this we decided to do research through design, meaning we would create a design, conduct user testing on said design, ideate solutions, then repeat the process with another design at a higher level of fidelity. In this way we progressed from conducting user research on paper sketches to a working web app prototype.
We began by conducting a competitive analysis on similar services and apps. Our initial idea involved creating a service that could generate archetypes of broad user segments, and manipulating the advertising users see. We found through our competitive analysis that similar systems already existed, and our industry mentor recommended we shift focus towards solving a specific problem.
After our competitive analysis my partner drew out low fidelity wireframes. We purposefully did not show each other our work until we had created a large number of designs so that we could generate as many differing solutions as possible without getting too fixated up on any one initial idea.
We then met up and discussed our ideas, refocused, took the valuable ideas we were able to generate, and created a low fidelity digital prototype in Figma. I performed remote user testing with stakeholders. We created a number of tasks based on typical scenario we might expect our users to perform, then had our stakeholders perform those tasks in our Figma prototype. I asked users to explain what they saw, understood, and were thinking as they progressed through our tasks. The feedback was invaluable and led to a total overhaul of our design.
We repeated this process of design and research with three additional designs. A medium quality Figma prototype, a high quality Figma prototype, and a working web app prototype. The web app prototype I built can generate text and images based on hard coded business for a platform you select. The prototype tests the novel areas of user interaction rather than the entire system, as our intent was to research User Experience and User Interfaces more so than an entire front and backend of a web app.
Early low fidelity black and white design
Mid-fidelity design
The timeline established