Lots of launches this week (with mixed success).
In the past week I've done some SEO work, some paid search, and some social stuff with mixed results. Lots more visits to the site but so far no conversions. Let's go topic-by-topic since that makes more sense to me than strictly chronological order
Natural language cron parser/generator
Two weeks ago, I released cron-naturally, an open-source typescript package which runs in the browser or in your preferred JS backend (node, deno, bun, etc). It translates the 5-position cron syntax into plain English, and most importantly, converts plain English into cron. I could not find any package that translated English into cron and I didn't want to include it in my product as an LLM call since that would be insanely slow and relatively expensive.
I felt like this was important because my product, Rills, can trigger workflow automations based on a webhook, an event in a third-party system, or a schedule. I'm targeting non-technical users so I wanted to build an easy-to-use experience where you can do an LLM-like thing of providing a schedule in English with helpful feedback if the input isn't possible (e.g. every 3 days). And because I didn't want to reinvent every wheel, I chose to power this off of cron on the backend which has limitations but is also incredibly powerful.
An alternative UI I built was similar to Google Calendar where you could specify recurring schedules by selecting days of the week or a dropdown of options for every N weeks on Wednesday, etc but that UX felt very clunky in my product so I wanted to simplify it.
Results were not exactly great. I launched to npm and am now seeing ~300 downloads per week, posted on Hacker News and got 1 upvote and 2 Github stars, and Google just started ranking the page in the ~40s but I'm hoping that'll improve. It's not exactly the target audience for my product but it felt like a good thing to do and I'm quite proud of it. I'll be doing more OSS releases as I validate what works well internally.
My first paid ads
I built a lead magnet for users of existing workflow automation platforms like Make and Zapier who want to add human-in-the-loop review which those platforms either make difficult or hide behind a paywall. The page provides an API key via email and it's limited to 100 reviews per month as a sample of one of the core features of my product. The difference is that Rills has a learning loop built-in which reduces your required reviews over time without increasing risk, so that'll end up in the marketing drip campaign.
With this new page, I also launched my first ad campaign on Google Search to attract leads looking for a safer way to build AI workflow automations. It's still too early to tell if I'll actually get any leads or if those leads will convert into customers but it's a cheap experiment and a genuinely useful tool that I'm hopeful will help people.
Continuing to manage SEO
Earlier this week I posted about my SEO journey and how I'm now back to levels I haven't seen since late April. I've made some additional improvements to capitalize on the extra impressions to get pages ranking a bit higher. Impressions are up over 200 per day now and continue to climb, but clicks are very low at about 4 per month. I'm now ranking pretty high on some valuable pages though so I adjusted some titles and descriptions to hopefully garner some clicks for the few pages in the position 2-5 range.
Some new pages have also joined the index so my average position across the site has dropped but I'm doing relatively well for some queries.
Social
Like I teased earlier this week, I started answering questions on Quora. I'm still doing it which has been both insightful and somewhat fruitful. I've gotten a small surge of traffic to the site as a result of the few answers where I added a link to a relevant blog post. I'm going to keep on doing this for now since I'm at a minimum learning more about how small business owners are thinking about AI, automation, and risk.
I'm busy though and have let Facebook, LinkedIn, Hashnode, etc drop off which is not good. I was starting to get some traffic from those sources which has helped with domain authority as well since some of my blog posts were picked up as sources on other people's blogs and ground.news. I plan on using Rills to automate this whole flow so it becomes easier for me to manage all of this posting while building out the sales funnel.
Wrapping up
I'm going to keep doing more of the same and see what clicks (pardon me for that one). I'm most hopeful about the paid ads right now since that seems like the most likely to have an effect quickly. While that's churning, I'm also going to keep submitting the product to all of the relevant directories as I get the time. So far that has felt like a complete waste of time though so I'm doing them sporadically so I don't get completely bummed out about it.
