Gibber
Head of Strategy & Product · 2021–2022
A B2B analytics platform built on Twitch and YouTube Live sentiment data. The tech was genuinely interesting. The sales cycle was not.
What I Learned
- Novel data is not the same as a clear buyer. You need both.
- B2B SaaS for marketers is hard when the buyer hasn't fully accepted that the channel you're measuring is worth measuring.
- Streaming was growing fast, but 'growing' and 'mature enough to have a budget line for analytics' are different things.
The thesis behind Gibber was straightforward: streaming platforms like Twitch and YouTube Live had become genuinely important media channels, but the measurement tools for marketers were primitive. You could see view counts. You couldn't see how the audience was actually responding -- what drove engagement, what triggered drop-off, how sentiment shifted across a broadcast.
We built a sentiment analysis layer that processed live chat data in real time and surfaced it as actionable signals for brand marketers and content publishers. The data was good. I still think the underlying idea was right.
The gap between "good idea" and "sellable product" turned out to be mostly about buyer readiness. The marketers who were most excited about the channel were often the most junior people in the room -- they didn't have budget authority. The people with budget authority weren't yet convinced that Twitch analytics deserved a line in their measurement stack alongside more established channels. We were essentially trying to sell a new measurement discipline into a category that was still proving itself.
We made progress -- some pilots, genuine interest from a few forward-looking brands -- but not enough to reach the scale that would justify the next raise.
Mixed outcome is honest. The tech worked. The market timing was probably 18 months early. I left having learned a lot about what it means to sell something into a category that doesn't have a budget home yet.