2
2 Comments

I grew search impressions 4.7× in a month. Almost nobody converted. Here's what I got wrong.

I grew search impressions 4.7× in a month. Almost nobody converted. Here's what I got wrong.


I built a budgeting app because I didn't like any of the ones I tried. Not in the "I could do this better" way founders usually mean — more that every one of them seemed to have decided how I was supposed to think about my money before it met me.

The categories worked one way. The rollovers worked one way. If your life didn't fit the rules, you bent your life to fit the app.

But the thing that actually made me start building was subtler. Every app wanted my bank login, and once transactions auto-imported, I stopped reading them. They were already there, already categorized. I'd glance at a dashboard, decide I was fine, and move on. That's not budgeting — that's bookkeeping with extra steps.

So I built the thing I wanted: zero-based budgeting, no bank sync, ever. You enter transactions — or paste a statement and let AI structure it — but you review every line. The friction is the product. It's the five seconds where you ask did I mean to buy this?

I've been shipping it publicly for a few months. Here's the honest scoreboard.

What worked: the content engine

I went hard on SEO, mostly comparison posts — "X vs Y," "best budgeting app for [persona]," "[competitor] alternative." Roughly forty posts.

It worked better than I expected:

  • Search impressions grew 4.7× in about a month (28-day window to 28-day window)
  • A new comparison post on an uncrowded matchup hit position 8 within 9 days
  • Blog engagement rate sits near 47%, average time on page just under two minutes
  • Organic search is now ~46% of blog traffic, roughly tied with direct

For a domain with essentially no authority, ranking page-one for commercial-intent queries in under two weeks felt like cheating. Content is the cheapest leverage a solo founder has.

What didn't work: everything after the impression

Here's the part nobody puts in their growth thread.

Click-through rate is 0.3%.

Not 3%. Zero point three. I rank around position 8 for my best cluster — genuinely page one — and the expected CTR at that position for those queries is about 2.4%. I'm getting a third of that.

Two reasons, and only one is fixable:

  1. "X vs Y" search results are saturated with AI Overviews. Google answers the comparison above the fold. The user gets their answer and never clicks. Even a perfect snippet at position 8 is fighting for scraps.
  2. My snippets weren't giving anyone a reason to click. I rewrote titles to lead with a verdict instead of a neutral description. Two rewrites in, CTR hasn't moved. I'm now fairly convinced the ceiling here is structural, not copy.

And then, below that: clicks that do land convert to signups at a rate I'd describe as aspirational. Blog-attributed signups over 90 days: zero.

Two things I got embarrassingly wrong

1. I counted spam as progress.

I'd been watching referring domains tick up and quietly feeling good about it. Then I actually audited them. Twelve of my thirteen referring domains were garbage — five pages from a PBN network literally titled "Boost your Google rankings with Premium PBN," anchored with text like "Buy Backlinks Online Cheap," and seven auto-generated "Website Stats" scrapers all resolving to a single IP in Moldova.

My real count of legitimate backlinks was one. A Reddit mirror.

Nothing malicious happened — scrapers scrape, and Google ignores them. But I'd been reading a spam counter as a growth metric for weeks.

2. I was optimizing the wrong end of the funnel.

I found a comparable indie budgeting app — same niche, same stage, similar content strategy. Ran the numbers side by side:

  • We rank for roughly the same number of keywords. Near parity.
  • They get about 100× the organic traffic.
  • They have 104 referring domains. I have ~1.

Same keyword surface area. Two orders of magnitude difference in traffic. The entire gap is authority.

Which reframed everything. I'd been treating this as a conversion problem — bad snippets, weak CTAs, unclear onboarding. It's mostly not. I'm ranking at positions 15–50 on keywords where they rank 2–11, and no amount of copywriting moves position 28 to position 3. I was A/B testing the doorway while the building had no street access.

The open question I actually want help with

Here's where I'd genuinely like input, because I keep going in circles.

My free tier is the whole point. Competitors charge $79–$109/year. My core loop — budgeting, transactions, debt payoff planning, goals, forecasting — is free forever, and I don't want to change that. It's the honest differentiator and it's why anyone tries this at all.

Paid unlocks three things: AI import volume, family sharing, and multiple budgets.

And here's the problem I didn't see coming: all three are late-stage needs.

Nobody signs up on day one wanting to share a budget with their spouse. Nobody needs a second budget for their kid in week one. Nobody hits an AI import limit before they've built the habit of importing. Every upgrade trigger I built fires months after signup — but most users decide whether this app is part of their life in the first week.

So I've engineered a paid tier that's genuinely valuable to someone who's been budgeting for six months, and completely irrelevant to everyone who just signed up.

Things I'm turning over:

  • Do I surface a "share with your partner" moment earlier — after a user's first balanced month — so the wall arrives with a real reason attached?
  • Should AI import be the hero of onboarding rather than a feature you discover later? It's the one gate tied to the app's core friction.
  • Or is this just premature, and the honest answer is that I don't have enough users for any conversion rate to mean anything yet?

I lean toward that last one, which is uncomfortable, because it means the work isn't clever funnel optimization — it's the slow, boring business of earning legitimate backlinks and waiting.

If you've built a freemium tool where the paid tier solves a problem users don't have yet: how did you create an upgrade moment that fires early without gating the thing that makes people love the product?

Genuinely asking. I'll take arguments that I've priced or gated this wrong, too.


Building BudgetLabs — zero-based budgeting, no bank sync, because the friction is the feature.

on July 30, 2026
  1. 2

    I enjoyed how much of this post is about separating evidence from interpretation.

    The same numbers can support several convincing stories, and I think that's one of the hardest parts of building—figuring out which story the evidence has actually earned you the right to believe.

  2. 2

    The line about optimizing the wrong end of the funnel hit me hard. I'm early stage too and keep catching myself doing the same thing — fixing the product when the real problem is nobody knows it exists yet.