Reforge

Would you trust an AI to interview your users? We did, here's what happened 👇 Reforge Insights just launched its new AI Interviewer feature. We weren’t sure if people would actually talk to an AI, so we put it to the test. Here’s how we validated it before considering it for regular use: 1️⃣ Invited a small group of users via our main dashboard to take an AI-led interview. 2️⃣ Offered an incentive so there was no downside. Even if they hated the AI, they got something out of it. 3️⃣ Asked the AI to collect feedback on itself at the end. The tool itself is kind of wild. You just describe in natural language what you want to learn. It builds out themes and questions automatically, and within minutes, you’ve got a ready-to-run interview. The feedback? Surprisingly positive. Most said it felt natural, and it did a solid job keeping the conversation flowing. It’s now part of our research toolkit, letting us gather qualitative insights at a much larger scale. Big shoutout to Chun Jiang and the Reforge Insights team for building something genuinely useful. (Link in comments if you want to give it a go.)

AI products need constant tuning. Updating prompts, adjusting model settings, token rate limits, and more. You’re no longer flipping a switch. You’re tuning a system. Most importantly, every adjustment can’t become an engineering ticket. Introducing Reforge Launch: AI-native feature flags and configuration. Launch lets you treat your product as a living system → You can test new models with specific customers. → Adjust prompts without deployments. → Roll out AI features gradually. → See what's working where. Reforge Launch gives product teams real-time control over the AI experience their customers see, while being designed for developers who expect their tools to be as intelligent as the AI features they're building. For developers, Launch is: → Agent-ready: AI coding assistants can automatically wrap new features in flags. → Self-cleaning: Automatically detect stale flags. → Custom objects: Store full objects such as prompts, model params, etc. → Type-safe SDKs for your stack: Native libraries for Python, Node.js, Ruby, Java, Go, and React. → MCP: MCP lets your coding environment talk directly to Launch Traditional software was deterministic. You shipped a feature, QA tested it, and it behaved the same way every day. AI broke this model. Launch is the fourth product in our suite for AI-native product teams. → We started with Reforge Insights to aggregate and analyze customer feedback. → Added Reforge Research to capture new insights through AI interviews and surveys. → Then recently released Reforge Build (beta) AI prototyping for product teams. Launch completes the cycle, enabling builders to go from insight to launch. Lots of information below.

Something worth sharing with Product folks: https://lnkd.in/grQhi-qY Brian Balfour article on Reforge covers multiple layers of what the product work with AI will really mean for us with great examples that summarizes the complexity of the software development new setup. Attaching my favorite here: Customers expectations on products relying on AI.

I always recommend Reforge for people who want to dig deeper into the mechanics of growing products. I am extremely excited to join Brian, Lauryn, Ethan, and Ely on the AI Growth Series starting this week. From a user acquisition perspective, Brian, Ely, and I have been discussing how AI is changing core growth loops, the main ways AI products acquire customers today, and the foundations needed to grow these types of products. 120K words (500 pages) worth of material across acquisition, retention, monetization, and defensibility. It's definitely not for folks that just want a few new tactics. Link in the comments to enroll 👇

I tried Reforge Build and found it truly fascinating! Highly recommend for anyone looking to enhance their product workflow with smart, context-aware prototyping.

Show, not tell. That's been my mantra when putting together the AI Productivity course for product managers. Too much content talks about the high level promise of AI or shows a vanity demo. Not enough shows you exactly how to use it to get real work done each day. That's why I'm proud that the course has 6 hours of hands-on demos, walking you through how to incorporate AI into your everyday PM workflows. The demos span: → Bolt, Magic Patterns, PostHog for prototyping → Claude, ChatGPT, NotebookLM, Whisper, Reforge for customer insights → Claude, MCP, SimplyPut, ChatGPT for data analytics → ChatGPT, Relay, Reforge for product strategy → Granola, Relay, ChatGPT for product execution If you've been struggling to gain leverage by incorporating AI into each and every deliverable you are responsible for as a PM, this course is for you. Don't miss out: This is the last week to join us before the cohort starts on Oct 14. Details and registration: https://lnkd.in/gVSPvWCT

Distribution channels are shifting. Product-market fit is being redefined. AI is changing how products are discovered, built, and used. When everything is crumbling, that's when it's time to go back to first principles and understand how growth really works to make sure you're prepared. I'm excited to be back teaching Growth Foundations with Reforge, a course on the core mechanics of retention, acquisition, monetization, and growth modeling for tech companies. These fundamentals remain timeless because they are rooted in user behavior and psychology of WHY and HOW people use products. They help you take stock, recalibrate, and adapt your strategy with clarity and confidence. I've got a great lineup of case studies again and we'll hear how these leaders are also adjusting to a new AI reality: Brent Zajaczkowski of Noom, Laura Schaffer of DigitalOcean, Richard Fleming (he/him) of Canva, and Holly Chen of Vizard. Ai. Enrollment is still open and sessions start on Weds 10/15. See you there (link in comments)