Original Summary

I've spent the last six months building an AI tool that redesigns existing PowerPoint presentations. After a lot of building, testing, and trying pretty much everything in the playbook to get customers, we're finally seeing some traction. The numbers are still small (okay, very small), but looking back it's interesting how wrong we were about almost everything. First mistake: building for the customer we wanted Our original plan was personalized sales presentations. Upload your branding, point the tool at a prospect's website, get a tailored pitch deck. Everything from our homepage to onboarding and pricing was written for sales reps. We spent about $2,000 on Meta and Google ads. Around 5% of clicks turned into free signups, but as far as we can tell, not a single one became a paying customer. In five months, we got four mid-tier subscriptions. All four canceled before their second payment. Not exactly the recurring revenue machine we'd imagined... Second mistake: pricing for the wrong behavior Everyone who paid had spent 1.5 to 2 hours inside the product first. People who made one presentation and left never paid. And why would they? Our free plan let them finish their deck and export it with a small badge. They'd already gotten what they came for. We had a $12 badge-removal option buried in Stripe that nobody ever used. In one 30-day period, we recorded exactly ONE checkout click. We'd built subscriptions for people who needed presentations regularly, but most users only needed to finish one. So we turned off ads and bet on organic At some point, paying to acquire users who weren't paying us stopped making sense. We decided we'd either figure out organic growth or move on. We'd read plenty about optimizing for LLMs, so we went all in on AEO/GEO, or whatever we're calling it this week. Schema markup, structured FAQ content, feature-specific landing pages, direct answers to questions people might ask an assistant, internal linking, Bing indexing. Basically every recommendation we could find. We weren't convinced it would work, but at least it didn't require another $2,000 in ads. Then something interesting happened. ChatGPT started sending us users We'd created a few pages about redesigning existing PowerPoints, a feature we'd barely emphasized. Google search volume for that problem was practically nonexistent. But those pages started showing up in ChatGPT answers. In one 30-day window: ChatGPT sent 53 visitors. About 35 signed up. Google organic sent 80 visitors. Zero signed up, mostly brand traffic and an unrelated free calculator. A 66% signup rate from ChatGPT referrals. Pretty crazy from someone who's spent a long time in SaaS marketing. We don't know which optimizations helped, but we'd stumbled into demand our keyword research hadn't revealed. So we doubled down on those pages, expanded the answers around that use case, and watched the traffic keep growing. And it

  • 情报分类：商业与市场研究
  • 分类依据：内容涉及商业、投资或市场动态
  • 信息来源：Reddit · SideProject
  • 发布时间：2026/10/15 05:53:10