Clarify

We just scaled our customer base 5x in two months! 🤯 Here's the crash course we got in managing customer feedback and why the "right time" to level up is earlier than you think. 🎯 Start with Slack Slack remains the gold standard for customer relationships. It's fast, timely, and with their new external connections feature, it's more powerful than ever. We prefix all customer channels with "external-" for clarity, while vendor, investor, and partner channels get their own distinct prefixes. Bonus: Aside from the organizational boon, this lets us see all the awesome custom emoji sets from our customers ✨ ⚡ Your old process will break at 10 customers This number might seem low, but it's where things got challenging for us. Why? Because managing customer communication isn't just about conversations. You need alerts when you're not glued to Slack. You need clear ownership of who's handling each request (or, as Harvard Business Review would say, "who's got the monkey?"). You need ways for engineers to dive in without drowning in notifications. And you need company-wide transparency without forcing everyone into every conversation. 🔄 Centralize all your customer requests For us, Thena was brilliant here. It creates a single channel to triage all customer requests. You control the sensitivity of what counts as a request. We've customized our workflow stages to match our exact needs: Needs Triage, Investigation, Resolving, Waiting on Customer, Waiting on Clarify, and Closed. 🎯 Update broadcasting is a game changer One of the Thena features we love the most is broadcasting. It lets us push updates to all Slack channels simultaneously. Instead of sending formal, stiff announcements, we can keep it personal. Patrick Thompson can share monthly product updates in his own voice, making them feel authentic and human. 💡 Log your key learnings religiously Your customer feedback system shapes how you understand and serve your users. For us at Clarify, it's doubly important because we're building a CRM. We need context about what people are asking for and how we're serving them to drive actionable next steps for our users. The right time to think about managing customer requests isn't when you're drowning in them—it's before that happens. Set up systems early, but keep them simple and scalable. What tools are you using to support customers in your early stages? I’d love to hear about them. And if you’re curious about Thena, shoot me a message and I’ll share our internal setup guide and integration!

Being creative with AI isn't a talent—it's a skill you build through deliberate practice. Here's what I've learned about making AI a powerful creative partner: 🛠️ Set yourself up for success: Make AI accessible everywhere. Pin it in your browser, add it to your sidebar, install the Raycast extension, and get the mobile app. The key is reducing friction. When AI is one click away, you'll use it more often and learn faster. 🧠 Rewire your habits: Whenever you reach for Google, try AI instead. Same goes for repetitive tasks like writing emails, filling in spreadsheets, doing research, and drafting content. Over time, you'll develop an intuition for what AI excels at and where it’s better to use another tool. 🎯 Think of AI as a focused collaborator: Ai is best at handling one clear task at a time. Don't dump ten requests into one prompt. Instead, give clear, specific instructions with relevant context and an example of what you want. Iterate from there. This is why "GPT columns" are so powerful—you can repeat the same prompt multiple times, tweaking just one variable to get different outputs. 💡 Pro tip: Break complex tasks into smaller chunks. Instead of asking AI to reinvent your entire workflow, start with one piece. Perfect it. Then move on to the next. The practices above have helped me do a *ton* more in the last 30 days, for way less effort. Here are some of the use cases I’ve seen the best results for: 📝 Generated email patterns for domain research 💓 Calculated threshold heart rates from workout data 🧠 Crafted witty website copy 🧑💻 Debugged command line errors 📅 Built a Raycast command for calendar management ⛰️ Solved API integration challenges The power of AI isn't in occasional brilliant insights—it's in the daily practice of finding new ways to leverage it to make your life easier. 🚀 If you’ve been practicing with AI lately, what has been your biggest lesson learned so far? For folks who haven’t meaningfully practiced with it yet, what repetitive task is the biggest time–or soul–suck on your plate today? Do you think you could hand it off to AI? Also - special shout out to Visual Electric for the image for this post. It's been a fun creative outlet to play around with their tool as well.

Product roadmaps shouldn't be treated like a to-do list. They're living documents that tell the story of your product's evolution and adapt with customer needs. 🌱 I learned this lesson the hard way. Early in my career, like many founders and product folks, I viewed roadmaps as rigid plans to execute against - plugging items into project management tools and checking off boxes each quarter. While this felt satisfying, this output-based approach missed the point entirely. 😬 Here’s the roadmap philosophy I live by today: The art of roadmapping isn't about the output - it's about the process. 🏗️ Or, as Winston Churchill offers, "Plans are of little importance, but planning is essential." When viewed through this lens, startup roadmaps are a tool that helps you: 🗣️ Articulate your vision clearly to different stakeholders 🎯 Prioritize work that moves you toward that vision 🛞 Adjust quickly based on customer feedback 🤝 Unite teams around shared goals For early-stage companies, I’ve found the best way to do this is to create and maintain three versions of the roadmap, each for a distinct audience: 1️⃣ Annual roadmaps for investors: Focus on major milestones and market opportunities 6-12 months out. This shows your strategic thinking and excites investors about long-term potential. 2️⃣ Quarterly roadmaps for customers: Share concrete value coming in the next 3-4 months. This builds excitement while maintaining flexibility to pivot based on feedback. 3️⃣ Internal roadmaps for teams - Break down the work into specific problems and experiments to tackle this quarter. This connects daily tasks to bigger goals (which we manage in Linear). For each version, the key is focusing on the problems you're solving rather than features you're building. This shift in mindset leads to better products, more engaged customers, and teams that understand the "why" behind their work. ❤️ The best roadmaps aren't about checking off feature boxes - they're about delivering value by aligning your team, customers, and investors around a shared vision for the future. If you’d like to dive deeper into my thoughts around this topic, I’ve written up a post on creating your first roadmaps as a startup founder on the Clarify blog. I’ll drop the link in the comments 👇

I often get asked how founders should define and measure success for an early-stage startup. 🌱 The truth is, there's no one-size-fits-all answer.  Having worked with thousands ( 🤯) of companies over the years, I've seen firsthand the challenges founders face in identifying the right metrics at each stage of growth. This question stayed at the top of my mind during my three years at Amplitude‎ after they acquired our previous startup and during my time working on growth at Atlassian‎.  After a bunch of rotations, I've started breaking down the success metrics to focus on based on the three early-stage phases of the startup journey:  💡Phase one - Pre-product: Focus on customer discovery. Your north star? 10 LOIs from ideal customers to kickstart your pilot. It's all about validating the problem, solution, and business model. 🌱 Phase two - Early Product-Market Fit: Invest heavily in customer success. Aim for $1M ARR within 12-18 months. Build your growth model and track acquisition, activation, engagement, and retention. 🌲Phase three - Growth: Scale operations, expand your market, and optimize efficiency. Target $10M ARR in under three years, following the T2D3 (Triple, Triple, Double, Double, Double) framework. But here's the biggest lesson when chasing early success: It's not just about the numbers.  Your metrics should tell a story about the value you're creating for customers and help you create a culture that people want to contribute to. 🚀 Afterall, "What gets measured gets managed." Don't lose sight of the qualitative aspects of your business. Balance your quantitative metrics with customer insights and adapt the framework to your unique situation. Building a startup is a marathon, not a sprint. 🏃 Persistence, adaptability, and learning from failures are crucial. Use these metrics as guideposts, not hard rules. What metrics have you found most valuable in your startup journey? I'd love to hear about what’s worked in your experience, and what metrics you’ve stopped obsessing over along the way.  If you want to dive deeper into my thoughts around measuring early success on the way to PMF, I wrote up an in depth post on the Clarify‎ blog that I’ll link in the comments. ⬇️  P.S. Shoutout to James Lee‎, Somrat Niyogi‎, and others for your feedback on the article. 🙏 ‎

ChatGPT helped me build the Clarify‎ Chrome extension and saved us 25-50% of the engineering time in the process. 🤯 I know that sounds hype-y, but hear me out.  I was super skeptical of how much you could actually use ChatGPT to support product development, but decided to give it a try at our team hackathon a few weeks back.  It gave me the ability to contribute meaningful code alongside our engineers and showed me how we can use ChatGPT as a tool moving forward to speed up prototyping.   Here’s how I approached developing our Chrome extension with the help of AI:   1️⃣ Scraping LinkedIn data: ChatGPT provided working code to extract profile information.  I used prompts like “It should scrape basic data from a Linked /in/ URL”.  2️⃣ Designing the UI: I uploaded a Figma mockup, and ChatGPT translated it into functional CSS. I leaned on ChatGPT to help me refine and style the interface with prompts like “Update the CSS of the button to match LinkedIn buttons” and “Style the popup to match the provided design.”  3️⃣ Debugging and refining: ChatGPT helped fix bugs and fine-tune API requests. This was where I spent the most time. I used prompts like “Remove the settings box and fetch the workspace slug from the API” and asked for help with questions like "How do I make the extension refresh only once on LinkedIn profile pages?" In hours, we had a working prototype. It wasn't perfect, but it was functional enough to validate the concept and get everyone onboard. From there, I tagged in John Jiang‎ from our engineering team to help make it production-ready by rebuilding the infrastructure, fixing bugs, introducing TypeScript and React, and adding continuous integration. This experience highlighted four main lessons for me: 1️⃣ AI can turn semi-technical operators into functional engineers. It's a game-changer for rapid prototyping and idea validation. 2️⃣ Create prototypes before committing engineering resources. This allows for faster iteration and better decision-making. 3️⃣ Engineers are still crucial for production-ready products. AI gets you far, but skilled developers are needed for polish and scalability. 4️⃣ AI can free your engineers from grunt work. Let them focus on high-impact features while AI handles lower-level tasks. For small teams, this approach is a force multiplier. It's not about replacing engineers, but about expanding capacity and accelerating development cycles. If you want more details on how I set up this project, check out my full write-up on the Clarify blog. I go into more detail about the development process, the prompts I used, and the lessons learned along the way. The link is in the comments 👇 

It’s Dreamforce week and the sheer size/power/influence of Salesforce is evident for everyone to see, so why would anyone dare to take on this behemoth in 2024? I recently chatted with Austin Hay, Co-Founder of Clarify, about his audacious move into the CRM space. His story might change how you think about the future of revenue tech. Here's why Austin decided to build a new CRM from the ground up: 1. The Integration Gap: After 10 years in MarTech and time at RAMP, Austin saw firsthand how existing CRMs struggle to unify marketing and sales data. The data models just weren't built for it. 2. The Developer Disconnect: "The world's best talent is writing in languages that are much more consumable like JavaScript and Python." Not Salesforce's Apex. Austin wants to build a CRM developers actually love. 3. The B2B2C Evolution: More companies are becoming B2B2C, but most CRMs aren't designed for this hybrid model. Austin saw an opportunity to build for the future. 4. The AI Reality Check: While everyone's talking AI, Austin's focused on practical automation. "We're actively not talking too much about AI because we don't want to overhype people." Actions speak louder than buzzwords. 5. The Operator's Frustration: Anyone in Silicon Valley has probably heard that nobody really loves their Salesforce instance, but it's not because Salesforce is a bad company or has a bad product. When people say that, they're expressing their frustration that things just haven't changed. Austin wants to build a CRM operators actually enjoy using. The CRM of tomorrow isn't just about better features. It's about: - Being an extension of your brain - Unifying MarTech and RevTech - Handling unstructured data from multiple channels - Automating next steps/actions, not just suggesting them One of my fave lines from my convo with Austin:  "The next wave of operators is coming and we just want to be prepared to help those people." Tune into my full conversation with Austin on the future of CRM on the latest episode of The GTM Podcast.