I'm 20 and my first SaaS flopped. Here's what I'm doing differently with the second one.
At the beginning of 2026, I built Clean Core, a focus and project management app for solo creators. I spent months on it. Built a whole website, wrote blog posts, planned a Kickstarter. Never got a single paying customer.
The problem wasn't the product. It was that I was building for a vague audience with a vague problem. "Productivity" is not a pain point. It's a category.
So I started over.
I looked at what I actually know, design, development, and enough about marketing to be dangerous, and asked what problem has a specific person in physical pain every single week.
The answer: marketing agencies logging into Meta, Google Ads, TikTok, and GA4 separately for every client, then spending hours building reports by hand.
So I built a dashboard that connects all four platforms, calculates blended ROAS automatically using GA4 as the source of truth, generates branded PDF client reports in one click, and sends real-time alerts when something goes wrong. I called it Clean Core Marketing Manager.
What I did differently this time:
Picked one specific customer: digital marketing agencies with 5-15 clients
Built around a real weekly pain: reporting overhead and platform switching
Launched fast instead of polishing forever. The app has been live for two weeks.
Did outreach before trying to grow. 17 cold emails to real agencies before any marketing spend.
I'm 20, studying Visual Communications at Ivy Tech in Indianapolis, and running Wonderworks Design Studio on the side. I have no funding, no co-founder, and genuinely need this to work.
The app is live at clean-core.app. Free forever plan, 14-day trial on paid tiers.
If you're an agency owner or know one, I'd be grateful if you tried it. If you've been through a failed first product, I'd love to hear how you turned it around.
The ICP discipline is the right frame. What also changed for us second time around: getting better information faster than competitors in the space. We automated regulatory monitoring with goffer.ai - tracks legislation through Congress by keyword and sponsor, fires Gmail labels when relevant bills move, SMS for votes. What was 45 min/day of manual checking is now 5 min reviewing the digest. That time went back into user conversations, which is where the real signal lives.
"Productivity is a category, not a pain point" — exactly. Most first-time founders (myself included) spend months building for the category instead of the pain. The shift to a specific weekly reporting overhead for agencies is the right move.
One thing that will make those cold emails and your landing page convert faster: instead of lead with the "connect all platforms" feature, lead with the Friday 6pm problem. The version where the agency owner is manually pulling data from 4 tabs while their client is waiting for the report is much more expensive than the "unification" version.
The Clean Core name worked for focus, but for agency analytics, it might feel a bit abstract. If this scales, moving towards a data-first brand would help.
I've been looking at agency-side conversion funnels this week — if you have a live demo or a specific report screenshot you're testing, I'd be happy to take a quick look and roast the above-the-fold clarity for you.
(Reference source: external_manual_ih_cleancorejack_fridaypain_may18_usd_presell_hv)
The Friday 6pm framing is genuinely something I hadn't thought about. I've been leading with the "everything in one place" angle which is exactly the feature-first trap you're describing. Going to rework the hero copy around that specific moment.
On the name, fair point. Clean Core made sense when it was a productivity/focus app, and I kept it when I pivoted because the brand already existed, but I've had the same thought. Not ready to rebrand yet but it's on the radar.
I'd actually love the landing page roast. I'll DM you a link to the current above-the-fold and one of the report screenshots I've been using in outreach. Fresh eyes on conversion clarity would be super useful right now.
you did the right thing, getting users before you launch. i tell most founders not to be in haste to launch, talk too your idea customers as many as possible. they will let you know if you are just wasting your time
100%. Talking to potential users before you're "ready" is something I had to force myself to do because my natural instinct is to keep building until everything is perfect. I could easily spend another three months polishing the UI and adding features nobody asked for. Had to actively cut myself off from that rabbit hole and just get it in front of real people. The feedback from those early conversations has already been more useful than anything I would have figured out alone in a code editor.
have you started getting decent number of users?
The shift you made is exactly right — "productivity" is a
category, not a pain. A pain is: "I spend 3 hours every
Friday manually pulling data from 4 platforms to build
a client report." You found that.
To answer your question at the end — I went through
something similar. My first attempt was also too broad.
What changed with the second one: I stopped thinking
about features and started asking "what does this person
do every week that they hate doing?" Then I built the
smallest thing that removes that one specific task.
For me it was indexing — SEO people waiting weeks for
Google to discover new pages. Annoying, repetitive,
completely solvable. That's IndexerPro .
10 days from idea to launch, first paying users in week
one. The specificity of the problem made everything
easier — the landing page wrote itself, the audience
was obvious, the value was immediate.
Good luck with the agency dashboard. The cold email
approach before any marketing spend is the right call.
"What does this person do every week that they hate doing" is a better brief than anything I've written for myself on this. The indexing problem is a perfect example of what that looks like in practice, super specific, obviously painful, and the value is immediate the first time it works. Landing page writing itself is exactly what I'm chasing with the reporting angle.
10 days to first paying users is the benchmark I'm trying to hit. Good to see it's actually possible when the problem is tight enough.
Good reflection.
The shift from “build a product” → “solve a specific recurring pain for a specific user” is usually the turning point.
Curious how early you validated demand before building the full version?
Honestly the validation was more market research than direct user conversations. I looked at what budget options existed for agency analytics and the gap was pretty clear, most tools in this space are either too expensive for smaller agencies or too basic to actually replace the manual workflow. That's where the pricing came from, competitive enough to be accessible but not so cheap it feels like a toy. There's also a free tier for life so agencies can get real value before committing to anything.
In hindsight I should have talked to more actual agency owners before building. That's the thing I'm doing now in reverse, which is backwards but at least I'm doing it.
The fact that you're building a second one after the first flopped already puts you ahead of 90% of people who quit after one try. That's not failure. That's tuition.
Most first SaaS products don't fail because the idea was bad. They fail because the builder never talked to enough users before writing code. If you're doing anything differently this time, the single highest-leverage change is validating the problem with real people before building the solution. Everything else — better tech, better design, better pricing — is secondary to whether anyone actually needs what you're making.
Building in public helps too. Share the journey. The audience you build during development becomes your first users at launch.
Curious — what's the biggest lesson the first one taught you?
What was the biggest lesson from the first one?
Curious whether it was a distribution problem or a product-market fit problem — those require very different fixes the second time around.
Honestly both. The marketing was nearly nonexistent, I was basically building in a vacuum and hoping people would find it. But I also think the product itself had a PMF problem at its core. A desktop app for general productivity is fighting uphill on two fronts: desktop is already a harder sell when everything has moved to the web, and "project management for solo creators" isn't something people are actively searching for a solution to. Nobody wakes up with that problem on their mind.
The combination of weak distribution and a product the market wasn't pulling for made it really hard to know which one was killing it. This time I at least want to make sure the pain is undeniable before I blame the marketing.
That's the hardest situation to diagnose — when both variables are broken at once you can't isolate the cause. "The market wasn't pulling for it" is the most honest PMF read I've seen from a post-mortem. Most people blame distribution first because it's easier to fix in theory.
Exactly. When product and distribution are both weak, every signal gets muddy. That’s why “the market wasn’t pulling for it” feels so clean. If there’s real demand, even imperfect distribution usually shows some kind of gravitational pull. When there’s none, blaming GTM can become a way to avoid the harder PMF truth.
The reframe you made is the one that matters: 'productivity is a category, not a pain point.' That distinction took most founders two or three failed products to figure out. You got there faster.
The 17 cold emails before marketing spend is the right instinct. One thing that will make those cold emails land better: lead with the pain you're solving before mentioning the product. The current pitch probably describes what the dashboard does (connects platforms, calculates blended ROAS, generates PDFs). The version that converts faster describes what the agency was doing at 6pm on a Friday before your product existed -- the manual export, the mismatched numbers from three platforms, the report that goes to the client tomorrow morning.
The agencies who respond to that version are the ones who have the problem. The ones who ignore it don't. This is more useful information than a landing page conversion rate.
On the GA4-as-source-of-truth point from the earlier comment: that's worth being explicit about in your onboarding. Most agencies are used to Meta and Google Ads reporting their own attributed revenue, and GA4 tends to show a lower number. Some clients get alarmed when the report they're used to seeing suddenly shows less revenue. If you pre-empt that in the first client setup, it goes from a bug-that-needs-explaining to a differentiation that builds trust: 'we show you the conservative truth, not the platform's self-reported number.'
17 cold emails is a tiny sample. What response rate did you get, and what happened when you followed up with the non-responders?
The Friday 6pm framing is something two people have now pointed out in this thread and I clearly need to actually implement it instead of just agreeing with it. You're right that the current pitch is describing the product, not the moment before the product existed.
The GA4 trust angle is a good reframe too. I've been treating the attribution gap as a disclaimer to get ahead of rather than a selling point. Positioning it as "we show you the conservative truth" is a much better setup for client onboarding.
Honest answer to your question: 17 sent, followed up with the non-responders, and still got nothing back. A few opens across both rounds but zero replies. So the emails are getting seen, they're just not compelling enough to get a response. This whole thread is making me think the cold email copy has the same problem as the landing page, describing the product instead of the pain.
Part of the issue is I'm still figuring out where my target customer is most concentrated. I know it's agencies, but which type, what size, and where they actually hang out online is something I'm still mapping out. That's probably contributing to the scatter.
The pivot you described is textbook — moving from 'productivity category' to 'agencies spend 4 hours a week on cross-platform reporting' is exactly the kind of specific pain that converts.
One thing worth thinking through as you scale: GA4 as source of truth for blended ROAS is a solid call, but GA4's data model is quite different from what most dashboards expect. Session-based vs event-based attribution can produce numbers that confuse clients when compared to Meta's own reporting. You may want to be upfront with agencies about that mismatch early — it's actually a feature (it shows the true view), not a bug, but it needs explaining.
I do a lot of work connecting analytics data layers for funded startups — if you ever run into weird DAX or SQL discrepancies when pulling GA4 data alongside ad platform data, feel free to ask. I wrote up a lot of the common DAX↔SQL conversion patterns here if useful: https://growthwithshehroz.gumroad.com/l/dax-to-sql-handbook
Solid execution for a 2-week-old product. Keep shipping.
The GA4 attribution mismatch is something I've already bumped into and you're right that it needs to be framed proactively rather than letting clients discover the discrepancy themselves. The "true view vs platform view" framing is actually how I've been explaining it in demos, good to know that's the right instinct.
Appreciate the resource, I'll take a look at the DAX to SQL patterns. That's exactly the kind of thing that's going to bite me as the data layer gets more complex.
Glad the framing resonates. The DAX patterns that tend to bite hardest are around time intelligence — CALCULATE with DATESINPERIOD behaves very differently from a SQL window function, especially when your GA4 data lands in UTC and your client's reporting timezone is something else. You'll also hit edge cases when blending GA4 session data with ad platform click data at the user level; the join keys rarely line up cleanly.
The handbook covers those scenarios with side-by-side translations. Good luck with the agency dashboard — the specificity of the problem you've picked is exactly right.
This is a much sharper second attempt because you moved from a broad “productivity” category into a painful agency workflow: client reporting, platform switching, blended ROAS, branded PDFs, and alerts when spend/performance breaks.
The strongest positioning is probably not “marketing manager” though. That can sound broad and generic. The specific pain is agency reporting intelligence: one place to turn scattered Meta, Google Ads, TikTok, and GA4 data into client-ready decisions and reports.
One thing I’d watch is the Clean Core name. It worked better for the focus/project-management version, but for agency analytics it may not immediately signal performance, reporting, or client trust. If this becomes a serious reporting layer for agencies, a cleaner SaaS/data brand like Beryxa .com would probably carry the new direction better than stretching the old product name into a different category.
The positioning feedback is genuinely useful and "agency reporting intelligence" is a tighter frame than what I've been using. Noted.
On the name, you're the second person to flag it in this thread so it's clearly something I need to sit with. Though I'll be honest, the very specific domain suggestion at the end made me laugh a little. Are you trying to help me or sell me something? Either way the underlying point stands, just had to call it out.
One practical thought here.
Since the name/category issue has already come up from more than one person, it may be worth pressure-testing the direction before you rebuild the landing page around the agency reporting intelligence angle.
I do focused naming and positioning audits for early products: current name risk, category framing, domain perception, whether the brand can carry the new direction, and what naming direction I’d take before more demos, client-facing PDFs, and public assets build around the current name.
Not a long consulting thing. Just a sharp written breakdown with practical recommendations.
I’m doing a few of these at $99 while refining the format.
If useful, I can do one for Clean Core and give you a clear outside read on whether the name can stretch into agency analytics, or whether it is better to separate the new product direction before it gets more baked in.
Best place to discuss privately:
https://www.linkedin.com/in/aryan-y-0163b0278/
Fair callout.
Both, honestly.
I gave the positioning feedback because the agency reporting intelligence angle is genuinely stronger than the broad “marketing manager” frame.
And yes, I mentioned Beryxa because it fits the direction better if Clean Core becomes a serious agency analytics/reporting layer rather than the original productivity product.
The bigger point is not “use this exact name or nothing.” It is that Clean Core now has category drag. It sounds like the older product idea, while the new direction is more data, performance, reporting, and client trust.
If you stay with Clean Core, I’d at least pressure-test whether agencies instantly understand the new category from the name. If they don’t, that friction compounds across landing copy, demos, referrals, and client-facing PDFs.
Beryxa is one clean option because it gives you room to build a broader SaaS/data brand without being tied to the old product frame.
I won’t turn the thread into a sales pitch. But if the name is seriously on your mind and you want to secure a stronger direction before the new positioning gets too baked in, message me privately and I’ll keep it straightforward.
For sure. Totally understand what you're saying. I will keep you in mind when I revisit everything. I will also shoot you a connect on LinkedIn just so we can stay in contact.
For those who have been through a failed product — what was the thing you changed that finally made the second one work?
This comment was deleted 2 months ago.