CampaignGPT Case Study

Case Study - CampaignGPT Hackathon Project

IZEA

SPRING

2023

AI BRIEFS & EMAILS

Context

Company AI Hackathon

Background

In early 2023, the launch of ChatGPT sparked widespread interest and excitement about the potential of generative AI. Our company decided to host an internal hackathon, challenging the team members to explore innovative applications using OpenAI’s API within the influencer marketing space. Participants were encouraged to build cutting-edge AI tools to streamline campaign workflow, enhance influencer discovery, boost campaign creativity, empower creator excellence, and more.

Goals

Build innovative marketing solutions leveraging the power GenAI

My Role

I developed a lightweight website using Bubble.io that seamlessly integrated OpenAI’s API to serve the needs of marketers. The tool provides campaign inspirations drawn from historically successful campaigns and streamlines the creation of detailed campaign briefs based on the input of the marketers. Additionally, CampaignGPT can draft of personalized emails based on the information and content from social handles, allowing marketers to quickly connect with influencers.

Time Line

2 Weeks in Design and Development

Drafting campaign

briefs

and

creator emails

time-consuming

task

Collaborative Campaign Brainstorming: An influencer marketing campaign begins with developing a campaign strategy that aligns with the brand’s vision, followed by creating a detailed brief outlining every requirement. They are looking for a system that understands their unique requirements, generating high-quality, customized briefs that capture the brand’s vision and strategic goals. This collaborative approach accelerates the ideation process, reducing manual effort while ensuring each campaign is both innovative and meticulously aligned with marketing objectives.

Time-Consuming Personalized Outreach: Crafting personalized outreach emails remains a labor-intensive process. Despite having detailed campaign briefs, the effort required to tailor each message to individual influencers is substantial. This manual personalization, from analyzing influencer data to drafting unique, compelling emails, often slows down the campaign rollout and diverts valuable time from strategic planning.

Process

How can AI solve this problem?

Campaign Strategy Recommendations

AI-driven recommendations can analyze campaigns from similar products and brands to inspire creative direction. The system curates innovative examples, enabling marketers to kickstart unique campaigns with fresh, data-backed creative insights.

Drafting Campaign Briefs

AI can streamline campaign brief creation by gathering essential data and translating it into an actionable plan based on the goals of the users. It can also collaborate with users to refine the brief based on their preferences, producing a focused and adaptable strategy.

Writing Personalized Emails

AI can personalize creator outreach emails by analyzing creator data to craft tailored messages. It can highlight key metrics and audience insights, ensuring each email communicates why the creator is a good fit to join the campaign.

Deliverable

Introducing CampaignGPT

CampaignGPT is a lightweight, Bubble.io-powered tool that integrates OpenAI’s API to provide campaign strategy inspirations, generate dynamic campaign brief recommendations, and draft personalized creator outreach emails. It increases marketer productivity by working with the users to produce quality campaign materials based on their needs.

step 1

Landing Page

On the landing page, it’s going to ask you for the brand name, the product you’re trying to promote, and the goal of the campaign.

Once the form is submitted, CampaignGPT

Asks OpenAI for campaign strategies

The prompt asks OpenAI to provide recent successful influencer marketing campaigns based on the brand, product and goal provided on the welcome screen.

Saves the campaign selections to database

The system parses the response into brand name, campaign name, description, and platform and saves the individual campaigns to the database.

Downloads brand logos from BrandFetch

Using a third party API provided BrandFetch, the system is able to get the brand logos for the campaign selection. It adds some visual to the next screen.

step 2

Recommendations

You can now see a curated list of past successful campaigns featuring brands and products in your vertical. Pick one you like and the system is going to draft your campaign brief and email based on your input.

After you select a campaign, CampaignGPT

Asks OepnAI to draft a campaign brief

The system sends a prompt to OpenAI asking it to write a campaign brief based on the sample campaign selected on this step. The brief would include an overview, a description, deliverables from influencers, and some influencer recommendations.

Asks OepnAI to write a campaign email

After the campaign brief is generated, the system asks OpenAI to come up with a campaign email template based on the brief.

step 3

Campaign Brief & Email Template

The custom campaign brief and email have been drafted by OpenAI, and they are written based on your brand, product, goal, and your campaign selection from the previous step. You can see the campaign brief on the left and a set of input fields on the right. You can fine-tune the campaign brief by providing tone of voice, # of words, or any additional requirements such as adding Instagram Reel as a deliverable or asking OpenAI to completely rewrite the campaign brief. You can even directly modify brief and ask OpenAI to generate a new version based on your edits. After you’re satisfied with your campaign brief, we will move on to the campaign email template. You can do the same with email template by inputting your requirements and asking OpenAI to rewrite the template for you.

step 4

Input Creators

After you're done tweaking the campaign brief and email template, you can input creators's information and ask CampaignGPT to generate personalized creator outreach emails.

Once you input influencers, CampaignGPT

Scrapes the creator’s Instagram profile with Apify

The system sends a request to Apify’s Instagram Profile Scraper and gets back creator information such as bio, latest posts, followers count, and many more data points.

Asks OpenAI to write the campaign email based on the creator’s information

The system feeds the creator information to OpenAI and asks it to write the campaign email based on the template while covering the following talking points: explain why the creator is a good fit, highlight the creator’s top performing content, and talk about how the creator’s audience resonates with the brand’s product.

step 5

Personaized Emails

Outreach Email Tailored to Each Creator

The final output of CampaignGPT is a uniquely tailored email that resonates with each influencer. It seamlessly integrates the creator's content and social handle information to craft messages that show how much you understand the creator, emphasizing the following key highlights:

Shows understanding towards the influencer’s niche
Highlights the influencer’s top-performing content
Expand market reach to attract users with varied marketing needs

After the emails are generated, you can review the emails and open them in your preferred email client to send the emails.

Feedback

The Result

Winning the Hackathon

I presented my project at our annual company retreat in Orlando alongside 10 other participants. Thanks to the support of my peers and the judging panel, my project was voted #1 and took home the grande prize of $5000. My tool’s innovative integration of OpenAI’s API and user-friendly design impressed peers and judges alike.

What Happened after the Hackathon

Following the hackathon, key design elements and the intuitive user flow of my project were incorporated into our platform, allowing marketers to easily generate briefs and personalized creator emails without leaving the platform.

My Learnings

Through this hackathon, I gained a deeper understanding of GenAI’s transformative power and the value of building AI-driven tools. The project provided hands-on experience with developing innovative solutions and helped me shape an AI-first product vision. In addition, building with Bubble.io and OpenAI's API allowed me to build with real data, rapid-prototype, and improve the quality of the user experience iteratively.

Work Samples

More Case Studies

Trackwise

Personal Project: Building Figma Plugin

Offer Workflow

IZEA Phase 3

See All Case Studies →