·7 min read

What 2,319 SaaS ideas reveal about 2026 — a data snapshot

Six weeks of scraping Hacker News, GitHub, Product Hunt, and Google Trends. Here's what the distribution actually looks like — categories, revenue tiers, difficulty, and the ideas that keep surfacing.


We run a scraper pipeline that pulls posts from four public platforms every day — Hacker News, GitHub trending, Product Hunt, and Google Trends — extracts SaaS-shaped product ideas from each one, deduplicates across sources, and writes them to a searchable directory.

As of this morning the database has 2,319 active ideas, extracted from 3,995 source posts. Every idea has a machine-scored difficulty, an estimated revenue ceiling parsed from a range string, and a one-paragraph commentary written by Claude Sonnet on market timing, competition, and biggest risk.

This post is the "what does that pile look like?" data dump. Every number below is a real count pulled from the production database on 2026-04-21.

Category distribution

14 categories. The distribution is extremely skewed.

RankCategoryIdeasShare
---:------:---:
1DevTools70230%
2Productivity46120%
3Creator Tools1968%
4AI/ML1788%
5Education1366%
6Fintech1286%
7Other1286%
8Health1004%
9Marketing854%
10Automation763%
11Ecommerce623%
12HR / Recruiting552%
13Logistics9<1%
14Real Estate6<1%

Half the catalog is DevTools or Productivity. That reflects where our scraper lives as much as it reflects reality — HN and GitHub Trending skew hard toward technical audiences. If you build for indie hackers, these are the two categories with the deepest pool of already-validated ideas and the most comp data per idea.

The bottom of the list is arguably more interesting than the top. Logistics has 9 ideas. Real Estate has 6. These verticals have real money and very few indie-hacker entrants because the discovery channels are off-Twitter. If you have domain expertise in either, the competitive density is an order of magnitude lower than AI/ML.

Revenue distribution

We parse the Claude-generated revenue range (e.g. "$2k-$10k/mo") into a generated Postgres column that stores the upper bound in USD. The distribution:

Revenue ceilingIdeasShare
------:---:
$50k+/mo522.2%
$25k-50k/mo441.9%
$10k-25k/mo62627%
$2k-10k/mo1,24354%
<$2k/mo1316%
Unknown / unparseable22610%

54% of the catalog is $2k-$10k/mo territory. That's the indie-hacker sweet spot: revenue high enough to meaningfully augment a salary or replace one, low enough that incumbents haven't fortified the niche. Filter to just this tier: $2k+/mo ideas.

The $10k-$25k/mo band (27%, 626 ideas) is where most "this is a business" territory sits — enough revenue to hire, enough comp to attract a second founder. If you're scoping toward a full-time quit, filter $10k+/mo.

The $50k+/mo ceiling is rare (2.2%) and usually reflects ideas with B2B enterprise angles that are hard to bootstrap solo. Browse them — even if you don't build one, they're useful as a reference for what "non-indie-hacker" SaaS looks like.

Difficulty distribution

Claude Haiku assigns each idea a difficulty 1-5 based on technical complexity and go-to-market difficulty. Mostly middle-of-the-range:

DifficultyIdeas
---:---:
1 (weekend)160
2 (easy)883
3 (medium)830
4 (hard)413
5 (very hard)36

75% of the catalog is difficulty 2 or 3 — buildable-in-a-month territory for a working solo dev. Only 1.5% of ideas rate as "very hard," and those almost exclusively involve building infrastructure (custom databases, specialized compilers, hardware) rather than applications.

The 160 difficulty-1 ideas are the most actionable subset for anyone looking to ship *something* this weekend. Filter to weekend builds to see them.

The ten most-mentioned ideas this cycle

"Mention count" is our proxy for demand signal. Each time the pipeline sees a post matching an existing idea (measured by pg_trgm similarity), the counter increments. The 10 highest right now, in rough order:

1. Pilot Flight Logbook Visualizer — 47 mentions. Small market, clear pain. Revenue: $300-1.5k/mo. 2. AI Codebase-to-Tutorial Generator — 32 mentions. Education category. Turns a repo into a learning path. 3. Novel Typing Practice — 31 mentions. Education, $300-1.5k/mo. 4. GitHub Issue Receipt Printer — 30 mentions. DevTools, novelty-ish, surprisingly $50-300/mo. 5. Developer-Focused AI Search Engine — 29 mentions. $5k-20k/mo ceiling. 6. Budget Bloomberg Terminal Alternative — 28 mentions. Fintech, $5k-20k/mo. 7. Deal-With-It Emoji Generator — 28 mentions. Creator tools, $100-500/mo. 8. Nugget — SaaS Idea Search Engine — 18 mentions. (Yes, a SaaS idea search engine appeared in our SaaS idea search engine. It's all recursion, no problems.) 9. Competitor Price Monitoring for Ecommerce — 17 mentions. $2k-10k/mo. 10. Blogosphere — Personal Blog Aggregator — 16 mentions. Creator tools.

Two patterns jump out. First, the top of the list is surprisingly niche — "pilot logbooks" and "novel typing practice" have nowhere near the audience of "analytics" or "CRM" but rank higher than generic tools because specificity beats scale in the data. People post about niche pains in detailed terms; generic pains get one-line complaints nobody scrapes. Second, several ideas at the top have modest revenue ceilings ($300-1.5k/mo). For a hobbyist or side-project builder, that's the signal — meaningful cash, manageable competition, real audience.

Patterns worth naming

Six trends that keep surfacing across the extracted data:

  • "Open-source alternative to X" is the single most consistent idea shape. Every popular SaaS eventually accumulates a self-hosted competitor in our catalog, usually within 18 months.
  • Subscription managers / cancellation helpers keep re-appearing independently. The wall between "I want to cancel" and "I actually cancelled" is a market nobody has closed.
  • Niche B2B SaaS (invoice reminders for tradespeople, equipment tracking for small manufacturers, scheduling for service businesses) is the most revenue-dense category per mention. Low mention counts (3-5), high revenue estimates ($5k-25k/mo). If you're hunting for undiscussed opportunities, these are it.
  • AI developer tools dominate sheer volume but have the most competition. Half the category is AI code review / AI testing / AI documentation variants.
  • Shared budgeting for couples has surfaced 5+ times from different directions. Specific enough to be real, general enough to scale.
  • Durable goods review platforms ("Rotten Tomatoes for appliances that last") is the most-mentioned consumer idea the internet keeps asking for but nobody ships.

Limitations of the data

Being honest about what this is and isn't:

  • Mention counts are weak-signal. A popular HN thread can drive 5 mentions in a week. The raw number isn't the same as "5 independent founders want this" — it might be one founder's post discussed 5 times. We mitigate by scoring mentions decay over time, so Week-1 hype fades by Week-4.
  • Revenue estimates come from an LLM. Haiku reads the source post plus context and picks a range. We've spot-checked hundreds against actual MRR data from public "$X MRR" threads and it's in the right order of magnitude ~80% of the time. Use it to rank, not to price.
  • Our source mix is biased. HN + GitHub + PH + Trends skews technical. Reddit and Indie Hackers are temporarily offline (upstream changes in the last two weeks; both coming back post-launch). The moment they're back, expect the Productivity / Creator Tools / HR / Fintech percentages to rise.
  • English only. Our extraction prompt assumes English discussion, which excludes a lot of specialized ecosystems.

What to do with this

If you came here looking for your next project, the highest-signal starting points are:

Every idea in the catalog comes with a category, difficulty, revenue tier, a one-paragraph commentary, and a free-to-use detail page that tells you why it's interesting and what's risky. No signup required to browse; ⭐ save the ones you want to come back to.

Browse all ideas → | Trending this week → | Highest revenue →