EarningsScores — Month 4 Update: 6 users, $0 MRR, and what happens when your site goes down for a month
Hey IH 👋
Building EarningsScores — an AI that scores every earnings report in real-time. Here's the honest update:
📊 Numbers:
- Users: 6 (flat from last month)
- MRR: $0 (still in free beta)
- Monthly costs: ~$200
- Tickers scored this month: 0 — site was paused
✅ What worked:
- SEO continues to be the only real acquisition channel (confirmed from Month 3). Organic search drove all 6 signups over the past 3 months.
❌ What didn't:
- The deployment got temporarily paused this month. Zero uptime means zero engagement and zero new users. This is probably the most honest failure post I've written.
🔧 What I shipped:
- Replied to IH comments from Nick06 and Marc about the $19.99/mo pricing question — their input shifted my thinking toward alerts/notifications as the paid feature
- Started scoping a "watchlist alerts" feature (get notified when a ticker you follow has a surprising earnings score)
💡 Biggest lesson:
A paused deployment is invisible until it isn't. I didn't notice the site was down for weeks. If you don't have uptime monitoring on your side project, set it up today — Betterstack has a free tier. Lesson learned the hard way.
📈 Next month's goal:
- Get the site back live
- Ship watchlist alerts MVP to the 6 existing users and ask if they'd pay for it
- Reach 10 users
Would love feedback from anyone in fintech or building AI products. And if you've dealt with Vercel/hosting pauses before — how do you handle cost management on a bootstrapped tool?
Trending on Indie Hackers
The watchlist alerts idea is probably the right paid wedge, but I’d be careful not to frame it as just “notifications.”
For someone following earnings, the value is not the alert itself. It is knowing which report needs attention before they waste time reading everything.
So the paid promise could be closer to:
“Get notified when a company you follow reports something unusually strong, weak, or surprising.”
That is sharper than “earnings alerts” because it sells judgment, not delivery.
Before building too much, I’d test this directly with the 6 users:
“Which 5 tickers do you care about, and would you pay $19/mo to be alerted only when something materially changes?”
If they do not care enough to name tickers, they probably will not pay yet.
The useful next step is not more SEO right now. It is turning those 6 users into a pricing/alert validation loop before trying to get 10 random new signups.
This reframe is exactly what I needed. I've been thinking about it as "alerts" because that's the feature I'd be building — but you're right that nobody wants to be alerted, they want to know when something actually matters.
The framing "get notified when a company you follow reports something unusually strong, weak, or surprising" is sharp. It sells the outcome, not the mechanism.
Going to email the 6 users this week with your exact question: which 5 tickers do you care about, and would you pay $19.99/mo to only hear from me when something materially changes? If they can't name tickers, that tells me everything.
Thanks for this — it's the most useful comment I've gotten on any of these updates.
Glad it helped.
The interesting part now is not whether the alert exists. It’s whether those conversations reveal a real paid behavior or just interest.
I wouldn’t try to map that out in the thread because the next few questions determine pricing, positioning, and whether this becomes a feature or a business.
If you'd like the tighter validation path, send me your email and I'll put it together properly.