Month Zero: Building a Marketing Engine for a $4 Todo App
The first update in our build-in-public series. Current MRR: enough for a coffee. Current strategy: content, social, and SEO for a joke product.
This is the first build-in-public update for anothertodo.app. I’ll be writing these monthly. The format is simple: here’s what exists, here’s what I tried, here’s what happened.
Let’s start at the beginning.
What This Is
AnotherTodo is a todo app. It costs $4 a month. It has five features: list, add, update, delete, toggle. That’s it.
It’s also an MCP server — meaning Claude can connect to it directly and manage your tasks via natural language. You tell Claude “add dentist appointment to my todo list” and it happens. The list shows up in the web dashboard, accessible from anywhere.
Is this revolutionary? No. Todoist exists. Things 3 exists. Apple Reminders exists and is free. The honest answer is that AnotherTodo is not trying to win on features.
Why It Exists
I’m an engineer who wanted to learn marketing.
Not “marketing” in the abstract sense. I mean the full stack: SEO, content, social media, paid acquisition, conversion rate optimization. The mechanics of how a product gets discovered and why people pay for it.
The problem with learning marketing is you need a product. And building a product you care about while simultaneously running marketing experiments is a nightmare — you end up doing neither well. So I built the simplest possible product I could ship in a weekend. A todo list with an MCP server.
The product is almost beside the point. It needs to be real — real Stripe integration, real authentication, a real thing you can pay for and use. But the interesting part isn’t the product. It’s whether you can build an audience and a distribution channel for something this mundane.
That’s the experiment.
The Strategy
There are three components.
Shitpost social. The product has a voice — deadpan, self-aware, slightly absurd. “Thought leadership from a $4 todo app.” The joke only works if you play it completely straight. No winking, no irony telegraphed in advance. You treat the product with the gravity of enterprise SaaS and let the comedy emerge from the gap. This voice is consistent across X, TikTok, and YouTube Shorts.
Deadpan SEO. This blog. Every post targets a real keyword — “build in public saas,” “mcp server tutorial,” “ai task management” — and the content is genuinely useful. The tone is still consistent. You can write a real, useful piece about AI productivity tooling and have it also be slightly funny. These aren’t mutually exclusive.
Build-in-public documentary. This post. Monthly updates on what’s actually happening — what’s working, what isn’t, what the numbers look like. Honest accounting. This is its own content format, and it targets people who are interested in the journey rather than the product. Some of those people will subscribe out of curiosity. Some will share the updates because they find the honesty useful. Some will just watch to see if it works.
What’s Been Built
Since the idea to now — call it six weeks — the following exists:
The product. A SvelteKit app, deployed on Vercel. Supabase for the database. Stripe for payments. One pricing tier: $4/mo. Subscribers get an API key and an MCP URL they can paste into Claude Desktop or claude.ai. The MCP server exposes five tools. The web dashboard shows your todos.
The blog. Built with mdsvex — markdown files rendered as SvelteKit routes. Per-page SEO with Open Graph tags and JSON-LD schema. A sitemap at /sitemap.xml. An RSS feed at /feed.xml. This post is the fifth one.
The content infrastructure. A backlog of post ideas targeting actual search terms. A TikTok/Shorts content plan. An X posting cadence.
The social presence. @alessiogrespi on X. TikTok account. YouTube channel. All posting under the same voice.
Marketing Channels: Why These, Why Not Others
The short answer: I picked channels I can produce content for without a team.
X (Twitter). Short-form text. Easy to produce, easy to iterate on voice. The build-in-public format has an established audience there — people who follow revenue updates, experiments, indie maker journeys. I’m not trying to go viral. I’m trying to build a consistent feed that attracts people who find this kind of project interesting.
TikTok and YouTube Shorts. Screen recording demos with deadpan voiceover. The content writes itself — “I added JSON-LD schema to a $4 todo app” is a 45-second video. This format also travels; people share oddly specific things if the framing is funny enough.
This blog. Long-form SEO content. Compound interest. A blog post that ranks doesn’t require you to keep feeding it. The return on a well-ranked post continues long after you’ve written it. The goal is to publish consistently and let the index build over time.
What I’m not doing: cold outbound. Emailing strangers about a todo list feels wrong. It’s also bad economics — the lifetime value of a $4/mo subscriber doesn’t justify the cost of manual outreach. If this scales, it scales through content and word of mouth.
What I’m not doing yet: paid acquisition. Maybe later, once there’s evidence that conversion works. Right now the conversion funnel isn’t proven. Spending money to drive traffic to an unproven page is a way to learn that your conversion rate is bad, at cost. Better to figure that out organically first.
Honest Numbers
Current MRR: enough to buy a coffee. One coffee. Maybe a medium.
Subscribers: single digits.
Blog traffic: just getting indexed. The posts are new, the domain is young, Google takes time.
X followers: growing slowly. The voice is landing — posts that play the joke straight get more engagement than posts that explain the joke.
This is normal for month zero. The point of publishing this isn’t to show impressive numbers. The point is that the numbers exist — something is live, something is tracking, something can grow. Month zero is infrastructure. Month one is distribution.
What’s Next
More content. The blog needs more posts. I have a backlog of SEO targets and I’ll be publishing 2-3 posts a month. Volume matters for organic search, especially on a young domain.
Consistent social output. The TikTok/Shorts strategy needs 10-15 videos before patterns emerge. What hooks work. What topics resonate. What makes people follow vs scroll past.
Conversion optimisation. The landing page is functional but not optimized. The copy is consistent with the voice but hasn’t been tested. At some point I’ll run variants and see what moves conversion.
Email capture. There’s no newsletter yet. There should probably be one — not for promotional email, but for monthly updates like this one delivered to people who opt in. That’s on the roadmap.
Next month’s update. In four weeks I’ll report back. Traffic, MRR, what’s moved, what hasn’t. The numbers will probably still be small. That’s fine. This is a long game and the interesting part is the trajectory, not the absolute values.
If you’re building something in public, the single most useful thing you can do is write down what you’re doing and why. Not for your audience — for yourself. Forces you to articulate the strategy, notice what you haven’t figured out yet, and commit to specific things you can be accountable to later.
This post is that, for this project.
Follow along at @alessiogrespi if you want the real-time version.