In the day to day at Arcade, I tend to very focused on sales and product. I see the formula for being a successful business as quite simple: build great product that solves people's needs, and sell it. I usually see activities outside of that as a nice indulgence. But last week, I did the opposite! Last week, I went to the Product Marketing Alliance conference in Seattle; co-hosted a product leader's dinner with Aditi Maliwal and had amazing people from Apple, OpenAI, Patreon, Gem, Gemini, Front, Faire, and Strike attend; and hosted an in person roundtable with our power users from Sentry, Retool and Carta. These are my takes after hearing some themes: AI came up over and over. This is not a surprise to anyone. If you are a product marketer or product manager looking to get insights from AI, you are better served to use the direct model over an AI wrapper. They evolve faster and are more customizable. But there is a lack of training on how to use the models. Surprisingly, not many companies are buying AI first tools. When I directly asked the leaders at these companies where their budget is going, I didn't hear as much spend on AI first tools. They are either building internal models or building on top of the models - or using tools that are solving a pain point with AI as a functionality. On the creator side, quality is a big issue with AI. Many people are using it to generate content and copy as a start, but they are not the final product. They are used "to get the ideas started" but, if presented with a higher quality option, they would quickly switch to that alternative. On the builder side, AI quality is also very hard to test. It seems to be from both a lack of sample data sets and effective training models. While the data exists, the permutations are so high that it's near impossible to predict every scenario and test for it. There's a risk with any AI launch to see how the market responds - and the quality only gets better with more adoption. That said, the pace at which AI models are evolving is almost shocking. They are near human parity, as seen from recent studies. Quality may not be a long term issue. Outside of AI -- Great products change rapidly, as well. As a tool that captures products at a point in time, it's a continuous challenge for our creators to keep them up to date. Asset search came up a fair amount. It seems like we're still at the early innings of companies being able to search across all internal content and knowledge. We got a lot of specific feedback about our product - and many of them mirrored our wish list for when we use Arcade ourselves. It was a nice reminder of two things: the simple things matter (down to how Arcade renders and looks) and to always use our product...or drink our own champagne 🥂 I failed to take a great photo, so here's an AI generated photo of our product dinner ;)
Tarik Remila
1mo
🔥
Alex W.
1mo
Great insights Caroline!
Founder and CEO at Pacific Tech Solutions
3wThe AI-generated photo adds a fun touch to your thoughts. Looking forward to seeing how Arcade keeps innovating!