How ForthWrite Uses Newsletter Placements to Grow a $0-1K SaaS

In my last finance role, I was sending 50-100 emails a day. I tried using ChatGPT and other AI tools to help, but I spent so much time rewriting their output, switching back and forth between tabs, and copy-pasting, that much of the benefit disappeared. Everything sounded formal, wordy, and robotic, and sending an obviously AI-written response to personal clients felt like reputational suicide. My inbox had gotten to the point where much of my day was spent triaging and replying to the highest priority responses, often letting follow-ups and other less important emails fall by the wayside. Ultimately, I fell so far behind that I started losing deals because of it, so I had to come up with a better way. That's when I started building ForthWrite, and initially, it was meant to solve that problem for myself, but I quickly realized that I wasn't the only one dealing with these challenges. Since then, we've gone to market to help others who find themselves in the same spot.

Though we're still trying to find the main acquisition channel, one that's worked surprisingly well is newsletter placements. Being well-regarded by a reputable brand seems to do well for transferring trust into a less well-known startup. Our biggest influx of new sign-ups came from the TAAFT newsletter specifically, but we've also had several other smaller engagement spikes from other mentions. As for ones that have flopped, there's been more than a few, but a notable one actually goes against the common consensus as far as launch milestones typically go, ProductHunt. We spent lots of effort rallying the troops to support the launch and got lots of surface-level attention but very little in the way of actual traction. My takeaway was that engagement and actual customer intent are very different things. If your target customers aren't already hanging out there, a respectable launch ranking can look much more meaningful than it actually is.

The biggest early friction has been setup and compatibility rather than the core writing experience. ForthWrite originally depended heavily on the Chrome extension, which meant users on unsupported browsers or locked-down work environments could hit a wall before experiencing the product. We've responded by making the web app more central, expanding supported surfaces, and continually simplifying onboarding. There are some blockers we're still currently trying to solve, such as IT firewalls that make certain Outlook clients much harder to gain access to. Because the individual often doesn't have the say to overrule these restrictions, we're building a top-down sales approach into our GTM funnel. By providing the tool to the company as a whole, we can work around those restrictions much more easily, as there now is an incentive to provide the access ForthWrite needs to work. It's like most AI tools in the sense that the more surfaces it's connected to, the more context it has, and the higher the quality the output will be.

I think the most common thing people get wrong about AI writing is assuming their communication is too nuanced or personal for AI to replicate. People will say that, “AI can’t write like me,” or “It doesn’t understand my job,” and with generic tools, they're often right. The thing is, that is largely a context problem, not a fundamental limitation of AI. If a system actually learns from how you communicate, understands the conversational context, and keeps track of the edits you're making to the generated drafts, the output can become remarkably close to what you would've written yourself. And if it gets something wrong, it will learn from the mistake and won't repeat it again in the future. I think the biggest hurdle for this category isn’t convincing people that AI can generate text. It’s earning enough trust that they believe it can represent them, and do it well.

The aim for ForthWrite is to be the communication layer for professionals so they can automate writing across all of the platforms they use. We're going to do this by expanding the voice-learning system across more communication surfaces, improving autonomous drafting, and building team functionality so individuals and companies can maintain a consistent voice wherever they write. Also, we're launching a consulting arm of ForthWrite, which focuses on AI implementation services for small to medium-sized businesses. The beachhead is financial service firms, as that's where my expertise is, but we'll likely expand to the broader market within the next 6-12 months.

We've hit 1,000+ installs on the Chrome Web Store with a 5.0-star rating and 5,777 drafts generated in product.

My relevant experience was with the problem we're solving, not the actual technical side of building. That started from scratch. The first iteration was an unlisted Chrome extension that required you to unpack the file folder in developer mode. It was an API call to OpenAI reformatted to insert into the reply appropriately. I learned by watching the co-founder at my last job build with tools like Cursor, Supabase, etc., and I co-opted some of his processes for myself. And honestly, I think I'm a good case study for anyone wanting to build their own product for exactly that reason. I hadn't written a single piece of code before starting ForthWrite; I just started with baby steps. You solve one problem, then the next presents itself, and you continue to build and add to your skill stack along the way, and before you know it, you have a business on your hands.

I fell into the common technical founder trap of building an extensive, deep product before finding the actual niche willing to pay for it. Because of that, the product is in a great place, but spending that kind of resource before PMF is likely a misallocation of scarce resources, especially early. If you don't know what people want, then how are you supposed to build it? Fortunately, the initial build solved a problem I was having myself, which makes me part of the ICP in some capacity. I trust my judgment, but I do not speak for everyone, and talking to more customers prior to building would've been the tactic I recommend to anyone else getting started. I think the hard part about this for most founders is that it requires collaboration and may result in answers that diverge from where you envision the product going. It's very easy to silo yourself off and build into your vision, but that may not actually be what's best for your product.