DEV Community: Aritomo Fukuda

X Banned Me During My Product Hunt Launch. Then Medium Saved Me.

Aritomo Fukuda — Tue, 28 Apr 2026 14:30:22 +0000

Twelve days ago, I launched OriginBrief on Product Hunt.

The launch went live, the comments started rolling in, and I did what any indie founder would do — I replied. To every single one.

Around 15 replies. Maybe 1–2 hours. Six of them included a link to my product.

That was apparently a crime.

The Suspension

The next morning, my @aritomofukuda account was frozen. No warning. No explanation. Just gone.

I filed an appeal. Crickets. I waited. Eleven days passed.

Eleven days during the most critical post-launch window — when comments need replies, when potential users have questions, when the algorithm starts deciding if you matter.

X’s spam detection had decided I was a bot. Because I replied to people too quickly. Because I shared my own product more than once. Because the engineers at X apparently can’t tell the difference between a startup founder and a spammer.

The Numbers Don’t Lie

Let me show you what I actually did:

Total tweets that day: ~15
Tweets containing a link: 6 (5 replies + 1 launch announcement)
Pace: roughly one every 6–15 minutes
Time window: 1–2 hours
Links pointed to: my own product (not a phishing site, not a sketchy affiliate, not a competitor’s page)

I’m a human being. I had just shipped something. I was excited. I was reaching out to my community on a launch day.

And X decided that was indistinguishable from spam.

What I Lost

For 11 days, I lost:

The ability to reply to my Product Hunt commenters
Connection with the indie hackers community I’d been building relationships in
Real-time growth signals that X provides during a launch
A meaningful portion of my expected post-launch reach

For an indie founder running on zero budget, that’s not a small loss. That’s the entire short-term growth lever, gone.

Download the Medium App

The Plot Twist

Here’s where it gets interesting.

Because X was dead, I had to do something I’d been avoiding: write longer-form content. So I started posting on Medium. Not every day, but consistently. Six articles in twelve days, mostly about the unglamorous reality of solo SaaS launches.

This week, I got my first “Read” — meaning someone actually read one of my articles to the end, not just clicked.

Today, the editorial team of Startup Stash sent me a private note. They have nearly 1,000 followers on Medium. They invited me to write for them.

Twelve days ago, I depended on X.

Today, I have a publication invitation and a relationship with editors who care about the work — not algorithms.

The Lesson

It’s not that X is evil. It’s that X is a platform you don’t control.

Their spam detection is a black box. Their appeal system is a void. Their support, post-Musk, has effectively stopped existing for individual creators. You can spend years building a following, and one day, an algorithm flips a bit and you’re gone.

The only sustainable channels for indie founders are the ones you own:

Your domain
Your email list
Your content on platforms that have functioning editorial relationships

Medium isn’t perfect. But when an editor at Startup Stash decides your work fits their audience, that’s a relationship — not a coin flip.

What I’m Doing Now

Still posting on Medium, now with publication amplification
Running the SEO content engine on my SaaS that auto-generates weekly reports (the original reason I built OriginBrief)
Leaving X to recover or not — at this point, it’s optional

If you’re building in public and X is your primary channel, please understand: you’re one false-positive away from losing it all. Build owned channels in parallel. Right now. Before you need them.

I learned this the hard way over the last 11 days.

Maybe this saves you from learning it the same way.

— -

I’m Aritomo, a 25-year engineer trying to figure out marketing as a solo founder. I’m building four SaaS products: StandupFlow, QuietLog, DocDecay, and OriginBrief. This is the kind of stuff I write about — usually with disclaimers about how little I know.

Disclaimer: I might be reading too much into 11 days of bad luck. Maybe X will unfreeze me tomorrow. Maybe Startup Stash will publish my piece and 3 people will read it. The point isn’t that I’m now successful — it’s that I’m no longer dependent on a single platform. That alone is worth the 11 days.

My SaaS Generates Its Own SEO Content. Automatically. Every Week.

Aritomo Fukuda — Tue, 28 Apr 2026 13:13:52 +0000

I have no marketing budget. My X account is frozen. I’ve written 5 Medium posts that have been read by approximately nobody.

So I did the only thing that made sense as an engineer: I made the product do the marketing for me.

The Problem With Content Marketing

Everyone says “create content.” Blog posts, newsletters, social media. All of it requires one thing I’m running low on: time.

I’m a solo founder running 4 products. I can’t write a new article every week on top of shipping features.

But here’s the thing — OriginBrief already generates content every week. AI research reports, from primary sources, on topics like AI industry trends, venture capital, cybersecurity.

For paying users.

What if I just… made some of it public?

The Build

I created a sample account with 5 research themes. Every Monday, OriginBrief automatically generates fresh reports for those themes — the same pipeline that runs for paying users.

Then I built /reports — a public page that displays those reports. No login required. Google can index it.

The result: every Monday, 5 new pages appear on my site. Each one is a structured research report with citations from real sources — company blogs, government sites, research institutions.

I didn’t write any of it. The product did.

Why This Works (In Theory)

Write on Medium
Google rewards pages that:

Get updated regularly ✓
Cite primary sources ✓
Cover specific topics in depth ✓
OriginBrief does all three, automatically, every week.

Over time, /reports/ai-industry-overview/2026-04-26, /reports/ai-industry-overview/2026-05-05, and so on — they stack up. Each one is a new indexed page. Each one is a potential entry point for someone searching for "AI industry updates" or "cybersecurity threats 2026."

It’s not a content strategy. It’s a content machine.

The Honest Part

Is it working? I have no idea yet. The pages went live 48 hours ago.

But the logic feels sound: if the product’s core function is generating research reports, the lowest-effort marketing move is to let it generate them publicly.

I’m not creating content. I’m exposing the exhaust.

The Takeaway

If you’re building a tool that produces output — reports, analyses, summaries, data — ask yourself: can any of that output be public?

Not all of it. Not the stuff paying users depend on. But the structural output, the kind that demonstrates what the product does — that can do double duty as SEO content.

Your product is already working. Let it work for you.

I’m sharing the journey of growing OriginBrief in public — a live SaaS that monitors primary sources and delivers weekly research reports.

I Almost Paid $1,250 for DR. Then I Found Out What “Dofollow” Means.

Aritomo Fukuda — Tue, 28 Apr 2026 13:09:37 +0000

Last week I almost signed a $1,250 contract for something I didn’t understand.

I’m a solo founder. I’ve been writing code for 25 years. I can debug a race condition in a distributed system, architect a multi-tenant SaaS from scratch, and ship four products in parallel.

But I didn’t know what “dofollow” meant.

The Pitch

It started with a LinkedIn message.

“Hey, I checked your site. Domain Rating 9. That’s below the threshold agencies care about. But I can fix it — DR30+ in 6 months for $700–1,250, depending on the package.”

I felt the sting immediately. DR9? That sounds bad. The way he said “below the threshold agencies care about” hit exactly right. He knew what he was doing.

He followed up with a menu:

DR Full Package: a 6–7 week project, $700–1,250 total (placement only, or full management with content writing)
Sponsored content potential: once you hit DR30+, you can charge $100–300 per article — a future revenue stream
And a kicker: “I can introduce you to clients once your DR is high enough. But we need to build the foundation first.”

Classic. Pay first, get value later.

I almost said yes.

The Doubt

Here’s the embarrassing part: I didn’t actually know what DR meant.

I’d seen “DR” and “DA” thrown around in indie hacker communities. I knew vaguely that it had something to do with SEO. But I’d spent 25 years on the engineering side — compilers, databases, distributed systems — and zero time thinking about backlinks.

So before replying, I did what any engineer would do.

I started investigating.

The Investigation

First: what is DR?

Domain Rating (DR) is Ahrefs’ score for how authoritative your domain is, based on how many quality sites link to you. It’s logarithmic — going from DR9 to DR30 is a much bigger jump than the numbers suggest.

Second: what’s a “dofollow” link?

This is where it got interesting.

There are two kinds of links on the web:

rel="dofollow" (or just a plain ) — passes "link juice," signals to search engines that the source vouches for the destination
rel="nofollow" — explicitly says "don't pass authority here"
For DR purposes, only dofollow links matter.

So I started auditing every directory I’d ever submitted to, looking for whether the link they gave me was actually worth anything.

I found traps everywhere.

Trap 1: rel=”nofollow”
Several high-DR directories listed me beautifully — and then stuck rel="nofollow" on every outbound link. The listing looks real. The link does nothing.

Trap 2: /aff/ redirect chains
Some directories route all their links through /aff/something?url=yoursite. This is an affiliate redirect chain. Search engines see the redirect, not the destination. No DR value.

Trap 3: elements
A few sites built their "visit site" action as a in a form, not an tag at all. You can't follow a button. Search engines can't either. Zero backlink value.

Trap 4: dofollow sold separately
Some directories offer a free listing (nofollow) and charge $79–99/year for the dofollow upgrade. They specifically build the product around monetizing the link itself.

Trap 5: API keys or secrets requested
One submission form asked for my Stripe restricted key. I had to stop and think.

I won’t pretend “always skip” is the right rule, because it isn’t. Established communities like Indie Hackers also ask for read-only keys to verify revenue, and I registered there. Reputation matters, and so does the scope of what’s being asked for.

The actual rule I landed on: judge by the site’s reputation and the permission scope of the key requested. An unknown directory asking for a live secret key? Skip immediately. A well-known community asking for a read-only restricted key? Evaluate the use case and decide.

When I finished writing this checklist, I opened my own site footer.

There Were 12

I counted. Twelve badges.

Download the Medium app
Not “without realizing it.” Every one of them was the result of me grinding through directory submissions during the chaos of launch week. I just hadn’t kept count, because the weeks around launch were brutal and I was running on fumes.

I checked each one.

Every. Single. One. was dofollow.

It wasn’t because I had a strategy. I’d just instinctively avoided the sites that felt off — the ones with too many ads, the ones that asked for weird things, the ones where the link looked wrong in the inspector. Twenty-five years of engineering had given me a pattern-matching instinct I didn’t know I was applying to SEO.

DR9 isn’t random. It’s twelve dofollow links from real directories, accumulated through six weeks of hand-to-hand work.

According to Ahrefs’ logarithmic scale, with five or six more high-DR listings — AlternativeTo (DR85), Alternative.me (DR74), Future Tools (DR69) — I’m looking at DR15–18 within a couple months. No paid campaign required.

The Decline

I wrote back.

“I appreciate the detailed breakdown. After digging into DR myself, I think I’m further along than I realized — I’ve got 12 dofollow listings already and a clear path to 15+ high-DR directories without paid placement. Budget-wise, I’d rather invest into product and content for now.”

He was gracious about it. No hard sell.

But I noticed something across the exchange: the “I’ll introduce you to clients” hook that appeared in the first message had quietly disappeared by the last one. The real product being sold was the DR service itself. The client introduction was the door-opener.

Knowing that made the whole conversation make sense.

One Thing Worth Saying

If you’re a solo technical founder who has been building in public for more than a month, you’re probably already doing more SEO work than you think.

Every product directory you submitted to. Every badge you added to your footer. Every “Featured on X” you put in your README. If you did the basic research — if you checked the link, if you skipped the sketchy ones — you’ve been building DR without realizing it.

The people who sell DR campaigns are selling the value of systematic application of what you may already be doing intuitively.

That’s worth something. But it’s not worth four figures when you can learn the five rules in an afternoon and do it yourself.

My checklist, for free:

Check for rel="nofollow" in the page source — if it's there, skip
Check if the link goes through /aff/ or similar redirect — if yes, skip
Check if “visit site” is a element — if yes, skip
Check if free listing gets nofollow and dofollow is paid separately — if yes, skip unless the DR is exceptional
If the site asks for API keys or secrets — judge by reputation and scope. Unknown site + live secret key = skip. Established community + read-only restricted key = evaluate the use case.

Six Months From Now

I’m at DR9 today. My rough projection:

AlternativeTo (DR85): +2–3 DR
Alternative.me (DR74): +1–2 DR
Future Tools (DR69): +1 DR
Dang AI (DR80, listed pending): +1–2 DR
3–4 more from my remaining list
Conservative estimate: DR15–18 by October. Without spending four figures.

Will I eventually pay for sponsored content? Probably, once I’m at DR40+ and the math makes sense for mutual value. But I’ll do it as a content partnership, not as a “please give me a backlink” transaction.

I almost paid $1,250 to learn a word.

Turns out “dofollow” is free.

I’m building OriginBrief — a continuous research platform that monitors primary sources and generates weekly intelligence reports. Follow along: originbrief.app

Previous posts in this series:

I Launched My SaaS on Product Hunt. A Week Later, I Still Have Zero Users.

Why I Stopped Reading the News (And Started Reading the Sources)

I’m a 25-Year Engineer Who Never Did Marketing. Here’s What My SaaS Launch Is Teaching Me.

Is Building in Public Overrated? Ask Me Again in a Year.

A note on this story: It’s based on a real exchange with one PR specialist over four days in April 2026. I compressed the timeline and paraphrased some messages and the menu of services to make it readable as a single narrative. The dofollow audit, the 12 footer badges, the five rules (including the nuance about Indie Hackers and other reputable exceptions), and the actual decision to walk away are all real.

Is Building in Public Overrated? Ask Me Again in a Year.

Aritomo Fukuda — Tue, 28 Apr 2026 13:02:30 +0000

The promise
I started writing about my SaaS launch before I launched it. Day counters. Screenshots. “I’m shipping X today.” The indie hacker playbook, pretty much verbatim.

The promise behind building in public is simple: you share the process, people get invested, and by launch day you have an audience rooting for you. A waitlist that converts. A community that shares your links.

That was the theory.

Nine days ago I launched on Product Hunt. I have zero paying users.

What the playbook said would happen

I followed the recipe as best a 25-year engineer with no marketing background could. Posted on X. Wrote reflections on Medium. Replied to other founders. Talked about what I was building before it existed. Showed the messy middle.

By launch day, I was supposed to have:

Some volume of inbound interest
A small group of supporters ready to upvote and comment
Enough credibility signal that strangers would trust the product enough to try it
Here’s what actually happened on launch day: I ended ranked somewhere around #40 with a handful of upvotes. Key Points score of 0, which on Product Hunt means you’re effectively invisible to anyone who doesn’t scroll to the bottom of the page.

Two days later, my X account got flagged and frozen. I submitted an appeal. I’m still waiting.

Three Medium posts in ten days. Total stats: 3 views, 0 reads, 2 claps from the same person.

What “building in public” quietly didn’t deliver

I don’t think building in public is a scam. I think it’s a good practice — sharing process, being honest about numbers, building a public record of your thinking.

Subscribe to the Medium newsletter
But the thing nobody says out loud is that the distribution part of building in public requires a different skill than building. You need to be someone who can write threads that get shared. Who can turn every small win into a narrative. Who is comfortable posting five times a day about themselves.

I’m none of those things. I’ve been writing code since I was a teenager. Writing code is a specific kind of output — it works or it doesn’t, and the output speaks for itself. Marketing output is the opposite. It doesn’t work until enough people decide it works.

Twenty-five years of training your brain to value “does the code run?” over “did the post get 2K likes?” is not easily undone in a launch week.

So the first lesson I’m actually taking from launching: building in public only works as advertised if you already have public-building skills. If you don’t, the “public” part never really activates, and you’re just building — quietly — with a blog attached.

What did happen — quietly

One thing did happen during launch week that wouldn’t have happened without it.

Someone from a country I’ve never been to — whose industry I’d never worked in — found my product on Indie Hackers and emailed me. Not from the launch-day rush. From a place where people still read past launch week.

He works in digital PR. I’d never thought about PR agencies as a user segment. He explained a concept called “newsjacking” to me, for free, across three emails. He wasn’t selling anything. He just saw the tool and saw a use case I hadn’t seen.

He’s still explaining concepts to me. No contract, no deal, no certainty it turns into anything. It might. It might not. But that conversation doesn’t exist if I hadn’t shipped.

That’s the honest version of “launching was worth it.” Not “users flooded in.” Not “went viral.” Just — one door opened that wasn’t open before. What’s behind it, I don’t know yet.

So, overrated?

Depends on what you were buying.

If you bought “build in public and launch day will fix your distribution problem” — yes, overrated. For most of us, it won’t. The loudest voices in build-in-public already had distribution before they started building in public. The causality runs the opposite direction from what it looks like.

If you bought “ship, write honestly, and over enough months some small number of useful things will happen that wouldn’t have otherwise” — probably not overrated. Just much slower than the Twitter threads suggest.

The post-launch phase I’m in now doesn’t have a 24-hour feedback loop. It has a months-long one. I won’t know if building in public worked for me until it’s been twelve months and I can count: how many emails like that one did I get? How many turned into something? How many people who saw a Medium post in April 2026 eventually trialed the product in September?

Ask me again in a year.

Until then, I’m going to keep shipping, keep writing, and try to stop measuring myself by launch-day numbers that were always going to be small for someone with my background.

The door opened. That’s all I know right now.

I’m a 25-Year Engineer Who Never Did Marketing. Here’s What My SaaS Launch Is Teaching Me.

Aritomo Fukuda — Tue, 28 Apr 2026 12:58:31 +0000

Shipping was familiar. Marketing is a separate craft — and I’m starting at zero.

For twenty-five years, I’ve been an engineer.

I’ve written code in a lot of languages, debugged systems at 2 AM, shipped features to production and watched them run for years. During all of that, I have done approximately zero marketing.

Last week, I launched my SaaS on Product Hunt. A week later, I still have zero paying users.

This isn’t a post about a comeback. I don’t have one yet. This is a post about everything I didn’t know I didn’t know, written from inside not knowing. If you’re an engineer who might one day sit where I’m sitting, maybe it’s useful.

The misunderstanding

For most of my career, I thought of marketing as something that happened after engineering. You build the thing, then “the marketing people” do their work, then users arrive.

The mental model looked like this: build → launch → grow.

I wasn’t dismissive of marketing. I just thought of it the way a software engineer might think of printer repair: important to somebody, definitely not me.

Now I’m doing all of it myself, and the mental model has quietly collapsed.

Marketing isn’t what happens after engineering. Marketing is what happens instead of engineering, for weeks at a time, when the product is already good enough and no one knows.

Five things I got wrong

Here’s what twenty-five years of engineering did not prepare me for.

I thought the product would speak for itself.

It won’t. A good product is a precondition for marketing to work — not a replacement for it. I’d heard this advice a hundred times. I hadn’t internalized it until I watched my Product Hunt page start at #41 early on launch day, drift into the 50s by the time the 24-hour window closed, and end with seven votes total — six from strangers, one from me. Zero comments. Meanwhile, products I thought were objectively messier got upvoted by friend groups I don’t have.

I built features when I should have written sentences.

I started building the product in earnest just over two weeks before launch. I didn’t start rewriting the landing page copy until the Monday of launch week.

A few days after going live, two different sources told me the same thing, independently. One was a founder I’d asked to look at the page. He said: “This copy reads like AI wrote it. Rewrite it.” The other was an AI-powered landing page review tool built by another founder — I’d run my page through it, and it came back with the same verdict.

An AI tool caught the AI-written copy. They were both right.

I’d spent just over two weeks on the product. A few days on the page that explains the product. Nobody was going to notice the two weeks of features. Everyone noticed the few days of copy.

I didn’t build an audience while I was building the product.

This is the most common advice in indie hacker circles. “Build in public.” I understood the words. I did not understand the compounding.

You don’t build an audience in the week of launch. You build it in the year before. Starting the month of launch is starting too late, and I started later than that.

I confused “submitting to directories” with “doing marketing.”

Become a Medium member
In the past week, I’ve submitted my SaaS to sixteen directories, counting the ones still pending. Some approved instantly. Some will publish me in May. Most will send me a small trickle of traffic.

Submitting to a directory is a box you tick. Doing marketing is a relationship you build. They are not the same thing, and I had them confused in my head for longer than I’d like to admit.

I treated launch day as marketing.

Launch day is the starting gun. It isn’t the race. I had spent so much energy preparing for the gun that I hadn’t really thought about the twenty-six miles after it.

Turns out, the twenty-six miles is the whole job.

What CS didn’t teach (but probably should)

Computer science taught me a lot. It taught me how to think about complexity, how to decompose problems, how to reason about correctness. It did not teach me any of this:

Marketing is empathy, compressed into a sentence. You aren’t explaining the product. You’re answering “why should this person care?” before they’ve asked. This is a skill with a learning curve — and I was starting on day one of the curve.
Distribution is a feature. The product with the best distribution will usually beat the product with the best architecture. This feels deeply unfair to engineers. It is also true, and rearranging yourself around it is part of the work.
Writing is a force multiplier. Every engineer I respect has opinions on variable names. Very few have opinions on the sentences that go on landing pages. Those are the same kind of skill, applied differently.
Positioning is a product decision, not a marketing decision. What you say your product is for changes what the product becomes. Positioning doesn’t happen after the build. It’s part of the build.
“Building in public” is a craft, not a hashtag. Sharing what you’re working on in a way that actually builds an audience is a skill. Posting screenshots into the void is not. I did the second one for a while and was confused when it didn’t work.
What I’m trying now

Here’s what I’m doing with the zero users, in case any of it is useful.

I’m writing. This essay is part of that. If I can’t run paid ads and I don’t have a big audience, the one thing I can do is leave behind honest, useful posts about what I’m learning. Maybe someone finds one in six months and finds me through it.

I’m talking to people. Not through Product Hunt comments — I got zero of those. Through emails. Through conversations that started on other platforms. Through every small interaction with someone who might have the problem my product solves. I’m not naturally good at this. I’m getting better.

I’m aware of the itch, but not acting on it. There’s a temptation, when you’re quiet, to add features that sound impressive in tweets. I feel it. Honestly, there are things I already know I want to fix. But right now, there are other things that need more attention, so those fixes are sitting in a list. I’m not resisting the itch — I’m just aware I can’t scratch it this week.

I’m accepting the timeline. Nothing I do this week will show up as revenue this week. That’s the part that’s hardest for an engineer. In engineering, you write the code, you run it, you see the result. In marketing, you write the code, you run it, and you see the result three months later, maybe.

A small note, from me to past-me

If you are an engineer who has spent twenty years, or twenty-five, or thirty, building things — and you’ve never really done marketing — here’s what I’d tell myself from a year ago:

You aren’t behind. You’re at the beginning. The beginning of anything is uncomfortable, and it’s especially uncomfortable right after spending a long time in the adjacent craft.

You aren’t going to be bad at marketing because of something you lack. You’re going to be bad at it because you’ve never done it. Though, to be honest, knowing this hasn’t translated into results on my end yet. I’ll let you know when it does.

Sit with the discomfort. Write a lot of sentences. Talk to a lot of people. Ship the product — and also ship the words about the product.

I’m a week in. I have zero users. I’m fine. I’m learning.

If you want to follow along, I’ll keep writing as I figure it out. If any of this resonated, you can try the tool I’ve been building at originbrief.app. Or just read the next post. Either one helps.

— Aritomo

Why I Stopped Reading the News (And Started Reading the Sources)

Aritomo Fukuda — Tue, 28 Apr 2026 12:54:55 +0000

Most “news” is a summary of a summary. Here’s what I found when I went back to the origin.

A few months ago, I caught myself reading the same story for the fourth time.

Same company. Same announcement. Four different articles, each a slight rearrangement of the same handful of sentences. I had spent twenty minutes reading four versions of one thing, and I still couldn’t remember the actual numbers.

That was the moment I realized: I wasn’t reading news. I was reading summaries of summaries.

The chain of whispers

Most of what we call “news” today is structured like a very polite game of telephone.

A company publishes a press release or a research paper. A reporter reads it, condenses it, adds a framing angle. An aggregator reads the reporter, condenses that further, adds a headline. A newsletter reads the aggregator, condenses that, adds emojis. A Twitter account reads the newsletter, condenses it to 280 characters, adds a take.

By the time the story reaches me, it has passed through four filters, each one stripping nuance and adding opinion. I’m not reading what happened. I’m reading what five different people, each with their own incentives, thought I should know about what happened.

This is not a conspiracy. Each link in the chain is doing honest work. The problem is the chain itself.

What happens when you go to the source

One evening, instead of opening my usual news feed, I decided to do an experiment. I picked three topics I cared about — a specific competitor, a regulatory area, a research field — and I went directly to the source.

I bookmarked the competitor’s blog and press page. I bookmarked the regulator’s announcements page. I found the journal where the research gets published.

The first thing I noticed: the primary sources are not hard to find. They’re just not optimized to be found. A press release doesn’t have a click-bait headline. A regulatory notice doesn’t have a hot take. A research paper buries the interesting finding on page eleven.

The second thing I noticed: they’re longer. A press release might be 800 words. A research paper is thousands. Reading the source takes five to ten times longer than reading the summary.

The third thing I noticed: it was worth it.

When I read the press release, I understood the actual announcement. The specific numbers. The specific product. The specific language the company used, which often revealed more than the language the reporters used.

When I read the research paper, I saw what the authors were actually confident about, versus what a headline had turned into a confident claim.

When I read the regulator’s notice, I saw what was binding, what was advisory, and what was still in comment period — distinctions that almost never survive a news article.

I was getting better information. I just wasn’t getting it often.

Become a Medium member
The problem with “primary source reading”

Here is the honest trade-off: reading primary sources is slow. It requires visiting different sites, checking different RSS feeds, learning which pages get updated and which don’t. Some press pages don’t have RSS. Some research journals lock everything behind paywalls. Some regulators publish on platforms built in 2003.

After a month of doing this manually, I had a clearer understanding of my topics than I’d had in years — and also, I had stopped doing it. The overhead was too high. The news feed, for all its flaws, was easier.

This is the gap I ended up building into.

What I built, and why I built it the way I did

I built OriginBrief because I wanted the quality of primary-source reading without the overhead of primary-source monitoring.

A few deliberate choices:

Weekly, not daily. Most “research tools” push updates constantly. But primary sources don’t move that fast, and the point of reading primary sources is that you’re not reacting to every tremor. One thoughtful weekly report beats fifty push notifications.

Cited, not summarized away. Every claim in the report links back to the source document. You can always verify. You can always read the original. The AI does the monitoring and structuring; the judgment stays with you.

Themes, not keywords. You don’t register “Tesla” or “autonomous vehicles.” You register a research theme — the thing you actually care about — and the system figures out which sources are relevant. A theme can have 5 or 15 sources, and you choose.

Focused, not exhaustive. The product only offers two plans: two themes or five. Not fifty. Because focus is the point. If you have fifty research themes, you don’t have research themes — you have a feed.

What I learned in the process

Building this taught me something I didn’t expect: the reason “news” feels bad isn’t that there’s too much of it. It’s that most of it isn’t actually telling me anything new.

When I started reading primary sources, the volume dropped dramatically. A week of real information was maybe ten or fifteen documents, not five hundred tweets. And yet I felt more informed, not less.

This is, I think, the real signal-to-noise problem. It’s not that the signal is hard to find. It’s that we’ve built an entire industry around packaging the noise to look like the signal.

A small reflection

I don’t think everyone should stop reading the news. I still read it — for breaking events, for context, for entertainment. I just don’t rely on it for understanding anymore.

For the topics I actually need to understand, I go to the sources. And I built a tool to do the going-to-sources part for me, because that’s the part I kept failing at.

If this resonates with you, you can try OriginBrief at originbrief.app. Starter is $33/month, with a 7-day trial. If it doesn’t, that’s fine too — I hope the rest of this essay was useful anyway.

Either way: the next time you find yourself reading the same story for the fourth time, consider going one link up the chain. You might be surprised what you find when you reach the origin.

— Aritomo

https://www.originbrief.app?utm_source=dev.to&utm_medium=blog&utm_campaign=news

I Launched My SaaS on Product Hunt. A Week Later, I Still Have Zero Users.

Aritomo Fukuda — Tue, 28 Apr 2026 12:49:04 +0000

Last Tuesday, I hit the launch button on Product Hunt.

A week has passed.

The user count is still zero.

Hi, I’m Aritomo. I’m a software engineer based in Japan, and for the past several months I’ve been building a SaaS called OriginBrief as a solo founder.

This is the part of the story that, in most Product Hunt recap articles, comes with a photo of champagne and a screenshot of a rising graph.

This article has neither.

What OriginBrief is

OriginBrief is a research monitoring tool. You register a topic — “my competitors,” “robotics regulation,” “quantum computing papers” — pick your sources, and it delivers a weekly report built from primary sources: official sites, research papers, press releases. Every claim is cited.

I built it because I spent too many evenings scrolling news feeds, only to realize none of it was actually news. Most of it was someone’s summary of someone else’s summary, optimized for clicks. I wanted something different.

That’s the product side. Now the honest side.

The quiet week

The launch itself was quiet. I didn’t have a big hunter. I didn’t have an audience on X. (Small side note: my X account has been frozen since April 17. That’s a separate story.)

Become a Medium member
Since launch, here’s what actually happened:

I submitted OriginBrief to about 15 directories. Some approved instantly. Some will publish me in late May. One said I could launch on October 19, 2026. I said thank you and closed the tab.
I found a bug where the initial report was invisible to every user. I discovered this before I had any users, which was, in a strange way, perfect timing.
I refactored two parts of the codebase, not because they were broken, but because I looked at them and felt sad.
Vercel’s dashboard informed me that my DNS settings were outdated. Apparently their IP range is expanding. I, somehow, was supposed to know this.
I had exactly zero sign-ups.
And yet — I’m fine.

What I’m actually writing this for

I noticed that most “launch recap” articles fall into one of two camps: “I made $50K in 30 days” or “I’m quitting to do something else.” Mine is neither.

I’m here, seven days post-launch, still building, still fixing, still learning. And I’d love to find the first few people who might actually benefit from what I built.

If any of this sounds like you:

You monitor a niche topic, but you’re drowning in tabs and newsletters.
You want primary-source research, automatically, not AI-summarized noise.
You’d rather get one good report a week than fifty mediocre emails a day.
…I’d love for you to try OriginBrief. Starter is $33/month, there’s a 7-day free trial, and if you email me, I’ll personally reply. I have the time.

What comes next

The graph will move eventually. Or it won’t, and I’ll learn something else. Either way, I plan to keep writing — a mix of build updates, things I’m figuring out, and opinions that might turn out to be wrong.

Thanks for reading this far. If you’d like to follow along, I’d be honored.

— Aritomo

P.S. If you’ve launched something recently, I’d love to hear how your first week went. Drop a comment. Zero judgment from me, especially.

https://www.originbrief.app?utm_source=dev.to&utm_medium=blog&utm_campaign=intro