Designing for creators
Flow diagrams for our viewer onboarding
Through our research, we understood that being a content creator is several jobs in one: creating content, managing fans, networking with other creators, working with sponsors, etc. To make matters worse, creators don’t have a lot of time on their hands. Creators typically support their content creation by working 9-5 jobs, leaving only a few hours outside of sleeping and eating to make great content.
Our initial goal with Twickle was to make a sustainable business model around a product that helped creators make a living. We set out to make a product that addressed the unmet needs of content creators, giving them the tools to automate their more tedious tasks and the freedom to focus on creating great content.
We started in 2017, designing different tools to help Twitch streamers. We had a few features under our belt, working with some content creators, getting some traction. We had a great connection to our users and released products fast. We had run some smoke test experiments that hinted at there being a successful business model with some of our features.
In late 2019, we decided to put it all together, go big, and go somewhat lean. We needed to monetize and grow Twickle into a high growth startup. Here’s the story of how we made Twickle a paid product.
The hordes of streamers flocking to our booth to click through the demo and chat (free user research!). Setting up and managing the booth was wildly exhausting and fun. We all lost our voices talking to streamers.
TwitchCon 2019
As TwitchCon2019 was approaching, we had a solid understanding of the needs of our users. We’d developed personas for Twitch streamers, Twitch viewers, and YouTube content creators. We decided that we’d continue designing for Twitch streamers since we worked most closely with them and the market size seemed big enough to serve as our beachhead.
At TwitchCon, we wanted to exhibit our aggregated toolkit that addressed the needs of Twitch streamers and get feedback from users on our vision. We had exhibited at TwitchCon before in 2018, but we felt this time our understanding of the content creation space was exponentially greater than that of before. Our partner program was in full swing and we were making a name for ourselves in the streaming world.
We came up with our grand vision of Twickle: a toolkit for everything a content creator needed to grow, engage, and monetize their audience. We diverged really wide in selecting features that would help streamers, looking at our personas and describing features to help with each pain point. We iterated on the features that would solve the needs of our users and got feedback early on from our partner program and research panel. This was a seriously audacious design that would help us understand what features were most important to streamers and help them see our vision.
Some swag we’d give out to our partners and new streamer friends. Everybody loved the purple fuzz.
Prioritizing features, creating an MVP, and testing hypotheses
At TwitchCon, we got 300+ signups and met invaluable industry connections. It was quite a time to be at Twickle. People loved the product and we had a solid user base to grow with. Most importantly, we were able to glean what the most important features were to our users from interviewing them and walking them through the prototype at the booth:
-
Giving fans points for supporting the streamer on social media
-
Giveaways to reward the fans who support the streamers most
-
Analytics to help streamers keep track of all of their content performance on different platforms
-
Fan manager to keep track of their fans and how they support on different platforms
We started looking like a customer relationship manager, designed for creators. People loved the product, but would they pay for it in a way that would scale us into a high growth startup?
To answer this, we decided to go forward and make an MVP based on our personas and feedback from TwitchCon.
We used a prioritization method from Laura Klein to focus on the first features to build, graphing our features on an XY axis. The X-axis measures how impactful the feature would be and the Y-axis measures how easy it was to build.
Prototyping our MVP
We were set on the minimum amount of work we’d need to complete this experiment. We then prototyped our MVP, pricing, and growth methods to see how users responded. Understanding how to price SaaS and how to grow is another topic altogether, but we learned a lot by sifting through best practices and chatting with founders of SaaS products.
We decided to opt for 3 different pricings, with a 7-day free trial for each. This model seemed to work best in our concept validation and usability tests.
(It takes awhile to load!)
Some screenshots of the product:
Streamers can see where their fans support them and how they last supported them
Aggregating all of the content creators analytics into one place. Magic!!
The streamer’s page, where fans would earn points and spend those points on giveaway entries
Testing hypotheses
After a month of concept validation, usability testing, and development, we were ready to begin setting up our metrics. How would we measure that this product is successful? We had 2 major hypotheses:
-
Value hypothesis: Providing actionable metrics and easier relationship management to the streamer will increase our retention
-
Growth hypothesis: We’ll find enough users through paid advertising and incentivized referrals to make a profit
Some of the more important metrics tied to these hypotheses:
-
Registration rate + funnel drop offs
-
Activation rate
-
Monthly recurring revenue
-
Churn (% of customers who unsubscribe)
-
Paid CAC (customer acquisition cost)
-
Referral rate
-
Average monthly value per customer
-
Unsubscribe reason (asked during the unsubscribe flow)
-
Retention cohorts by day
Aside from that, we would directly contact our users to interview them to triangulate our quantitative data.
Up to this point, we had over 5,000 users and were growing. Morale was high, we felt good about our product and our direction. What could go wrong?
Things went horribly wrong (not really)
We launched our MVP in February 2020, with Stripe, HotJar, and Mixpanel baked into our software, ready to measure. We used Facebook ads, Twitter ads, incentivized referrals, cold messaging, and sponsorships to get traction.
Through our measurements and interviews with users (paid beta users, unsuccessful conversions, and users who unsubscribed), we learned:
-
Some users will pay. These were our wonderful, bold, lead customers.
-
Not enough users would pay. We likely wouldn’t be able to make a startup growth company with these metrics if we decide to scale.
-
Having users enter their card before trialing the product was a huge hindrance in onboarding. We followed the advice of adding more friction to onboarding to make sure only lead users make it through, which was the wrong tactic for our market.
-
Some creators just don’t want to pay on principle. They believe software products ought to be free for streamers.
-
A 7-day trial is not long enough. Streamers needed more time to integrate Twickle into their flow to see the benefits.
-
We need to find a better traction model. Referrals and ads were okay, but we’d likely need a partnership with a larger company servicing this market to get the right traction.
To make matters worse, the COVID-19 pandemic had begun sweeping the world.
We decided to reduce the burden on our users and make Twickle free to use for the duration of the pandemic. They are already a heavily impacted population and we wanted to support them in any way we could.
We still had some lingering questions about our experiment. There were so many variables to tweak at each step of the customer journey (e.g. maybe we do a free trial without having them enter a card first). We could have set up hypotheses for each one and run different tests on them. A small product pivot could probably increase our conversion rate.
However, our gut feeling was that this wasn’t the right way to monetize in a way that’d make a profitable startup.
Finding a new user in the content creation universe
This entire product was an exercise in lean startup methodology. The experiment had proved this wasn’t a viable product hypothesis. We needed to pivot.
All along our 400+ interviews, we had learned about esports teams as a possible new market. Some streamers we knew were involved on esports teams and we had met some connections at TwitchCon 2018 and 2019. We then decided to keep our creator product free and explore this new esports terrain, marking another test in our lean startup journey.