Swayware: Fan Engagement Platform

Platform for sports teams to personalize fan experiences, from merch to media.

Comment

Inspiration

​​Non-traditionally for most hackers, Emory comes from a go-to-market background. He led sales teams at numerous AI companies while teaching himself code, and understands the challenges brands face in engaging consumers. Tejas worked as a Senior SWE at Expedia, and knows the technical challenges involved in serving consumers with effective experiences. We have a company based around providing personalized consumer experiences, and

When Kirk mentioned the Warriors goals of becoming a media company, it confirmed what we wanted to build was valuable, and can help them reach their goals of becoming a more global brand closely connected to their fans.

What it does

Swayware Fan Engagement Platform allows sports teams to understand who their fans are at an individual level, predict what they will do next, and personalize their commerce experiences 1:1 all in a single solution. We can change what each fan sees, give them discounts, and even tell the Warriors how to get them to engage more with the brand.

We collect anonymous clickstream from fans engaging on Warriors digital properties and using a proprietary (patent-pending) algorithm called Zero-Shot ID that predicts their identity and behavior, and what actions they’re likely to take next. We give these insights back to the Warriors so they can get value and take action on them.These insights can be used to target consumers with Swayware plug and play marketplace solutions like image personalization and dynamic pricing.

How we built it

Data Collection and Warehousing

On signing up to Swayware Fan Engagement Platform, you get a script which you can insert on your sports website. This script tracks anonymous visitors (powered by Snowplow) and the collected event data is pushed by Snowplow into Swayware’s Snowflake warehouse instance.

Running ETL

We run ETL jobs on warehouse data to produce output with proprietary (patent pending) algorithms that give insights into fans’ identification and behavioral data. Although this is not fully automated, we use open source ETL software to run these workflows.

SaaS platform for Fan Engagement

We’ve deployed the SaaS platform to AWS and is accessible from app.swayware.com. This powers customers (sports and e-commerce companies) facing workflows. Here our customers can see how the merch or products they sell are performing and insights into their fans/consumers like predicted age-range, predicted education level, behavioral traits like buying intent, discount seeking, color preferences and so on. The tech stack is primarily AWS Serverless postgres, Next.js, Prisma (ORM) and Tailwind.

Marketplace solutions

Swayware Dynamic Pricing marketplace solution offers connecting the prediction insights to your website directly where we can update the product prices based on the anonymous visitor’s behavior. We also have an Image Optimization solution live on Shopify used by tens of e-commerce companies, and we have leveraged a similar tech stack to build a dynamic pricing solution.

Challenges we ran into

We could not deploy the ETL workflows to the cloud within the hackathon time-frame and had to do it on a local machine. In the near future we want to prioritize this deployment so we can serve our customers way faster.

Accomplishments that we’re proud of

Creating a comprehensive platform for the Sports industry and focusing on fan engagement, and presenting gathered insights via easy to understand UI/UX.

What we learned

Understanding the sports industry and what matters the most to them, bridging the gap between gathered behavioral data and applying it for fan engagement.

What’s next for Swayware: Fan Engagement Platform

We will be rolling out the capabilities we built to our existing customer base, and our new partners. We would love to have the Warriors give the fan data feature a try, and see what information about their fans we can uncover for them. Setup takes about 60 seconds, just takes copying and pasting 1 line of code.

Built With

Share this project:

Updates