TSC | Previously The Starr Conspiracy

Here is a low key post about a solution we started moving to production this weekend. Zora is an agentic marketing platform. I know that sounds like hype, but that’s okay. Basically, Zora delivers what marketing leaders really want from AI … an autonomous solution that generates quality, on-brand content and automates many high-value marketing activities (like competitive intelligence, social media management and outbound marketing). There are (at least) two big things that make Zora better than anything you are using now. First, it automates the creation of fine-tuned LLMs. When using a generic LLM (like chatGPT), generating distinctive content is impossible. But fine tuning is currently mysterious and difficult. Zora will fine tune a model for you. All you have to do is upload content and walk away. When the fine tuning is over, you have a model that has your voice baked in (and your knowledge). Second, Zora is a multi-agent platform. Agents are very specialized AI workers that collaborate in groups. One agent might be a researcher, another might be a writer and yet another might be a publisher. And they can work together collaboratively to achieve objectives created by the user. Today was a big day. Developers created Zora’s first agent. This agent only has one job. Turn documents (uploaded by the user) into pure, unformatted content. After the user uploads a document, this agent stores the document in Google Cloud Storage, writes the metadata to a MongoDB collection and writes the clean, unformatted content to yet another MongoDB collection. Pretty boring. But it’s a huge step in the fine tuning process. It’s going to take about a dozen different agents working together to generate a fine-tuned model. But when all the agents are built, users can get their own fine-tuned model In a matter of hours without caring about how any of it works. And once you have a fine-tuned model? Everything else kinda falls into place. Before the end of the year, Zora will be available to clients of TSC. Stay tuned for updates on progress. But don’t tell anybody.

Measurement killed the marketing star. I resonated with a post on this topic from Tas Bober so much that I asked her to join me on Episode 70 of Micro Podcast to discuss it. She didn't disappoint! Tas had this take: - There used to be a time in marketing when creativity was king - Long-term branding building was the goal. - There was magic in the unknown But then along came digital advertising Which was followed by marketing automation. It became much easier to measure and track the effectiveness of our marketing campaigns through digital channels. And our industry got addicted to attribution and real-time data. If we allocated a dollar toward a campaign, we wanted to understand which deals we could tie back to that investment. If we couldn't assign direct attribution, then that campaign got killed, even if we knew it was good work and would resonate with our ICP. And we allocated those funds to areas where we could prove more of a direct attribution (cough..cough...Google). Then we focused on how we could make incremental improvements and optimizations. This was how we tried to beat the competition. I think there was a belief that we could take the risk out of marketing All we had to do was measure everything and only invest in what we could prove had direct attribution. As Tas says in her post: "We shifted from being creatives to scientists." Then, that whole model broke in 2023. So what now? It doesn't mean that we shouldn't invest in demand generation. But it does mean.... -Marketers must get back to using creativity to connect with their buyers -We have to make peace that we just can't measure everything -Investing in brand may not have an immediate impact on this month's pipeline... -But it will absolutely determine our success in the long term What do you think? JJ Lance Steve Bret Ashley Bill Hailey Anthony Shannon Julie Noah Jonathan Jonathan John Chris Sam Be sure to check out Tas' original post. Link in the comments below ??