Coining a Term AND Creating a Category

Creating a new category is not for the faint of heart. It requires patience, persistence and substance to make it stick. Just ask Greg Kogan, VP of Marketing and Growth at Pinecone, the leader in vector databases.

When Greg first joined Pinecone as a consultant, the term "vector database" didn't even exist. Greg and the founder spent months having late night conversations, interviewing target users, and testing different category names before landing on "vector database" (the other leading option was Vector Search API). Even then, it took many months before the terminology started resonating with people.

"Nobody gets it. But there's some sort of intrigue. It's getting attention. So let's stick with it," Greg recalls thinking at the time. He estimates it took a full year before "vector database" was embraced by customers, competitors and analysts. Fast forward to today, and the term "vector database" has become table stakes in the machine learning infrastructure world with over 100 vendors claiming to offer a vector databases.

Greg advises marketers who are considering category creation to keep an open mind, look for early signals that the messaging is resonating, and have realistic expectations that it could be a long journey. Most importantly, avoid creating a category just for the sake of differentiation.

From 0 to 10,000 Signups a Day via PLG

In addition to branding a new category, Pinecone has also been on the forefront of the product-led growth (PLG) motion in the AI/ML space. The company went from 0 to 10,000 signups a day at the peak of the generative AI hype cycle in 2022.

Pinecone was able to gradually build and scale their GTM engine through experimentation - hosting webinars, attending meetups, and most importantly, creating a robust self-serve product experience. Some key initiatives:

• Streamlining onboarding by removing unnecessary features that created friction
• Investing in high-quality documentation, with the docs team sitting in Growth
• Defining key use cases and target personas to focus GTM efforts  
• Building a customer support chatbot using their own product
• Partnering with consultancies and service providers for implementation

For companies embarking on a PLG journey, Greg recommends closely tracking how users are engaging with the product, getting their qualitative feedback, and making iterations quickly. He also suggests being selective about target audiences based on their likelihood to convert.

"I talked to a lot of people in that first year. I was out in the community myself, listening intently on Hacker News, and setting up Google Alerts for relevant keywords to see what people were talking about on the topic," he shares. While Pinecone has been in the thick of the generative AI revolution, Greg doesn't believe AI-native companies necessarily have an advantage when it comes to marketing. Instead, he encourages all marketers to focus on customer outcomes, regardless of the underlying technology. "A year from now, everything will have AI underneath it. And so it's going to be back to marketing it like any other product...it puts the onus on us, on marketers, to really understand what our audience's desired outcomes are and speak to that," he says.

Developer Advocates are Key When Marketing Dev Tools

When marketing to developers, it's critical to lead with substance and education rather than hype and jargon. That's why one of the first go-to-market hires at Pinecone was a developer advocate. Their first Marketing hire was a dev advocate who was an ML engineer himself before joining Pinecone. You have developers teaching developers how to build amazing things - what's not to love there? While the dev advocate role sits under the Growth team at Pinecone, their primary focus is on creating helpful content, tutorials and sample code, not pitching the product. The goal is to genuinely help developers solve problems and explore interesting use cases.

Companies like Stripe and Twilio have shown how powerful this community-building approach can be.When scouting dev advocate candidates, Greg looks for two key qualities beyond just technical chops: "an inherent desire to teach people cool things" and strong communication skills. Avoid hiring someone who is purely a subject matter expert, he cautions. "

The key lessons? When marketing a highly technical product to a technical audience, lead with substance over style. Hire practitioners who can speak the same language as your users. Have the conviction to create a new category, but balance it with humility and a willingness to iterate based on feedback.