Simon Høiberg: From $300 MRR Team Mistake to a $100K+/mo Portfolio
Danish bootstrapped founder who built FeedHive, Aidbase and a multi-product stack, learned to stay tiny, and now runs a high-revenue SaaS portfolio with heavy AI automation.
Editorial profile. Written by IndieFounder Editorial Team from publicly available information. It is not a direct interview: the first-person answers are our summary of the founder's public statements, not verbatim quotes. Figures are approximate and not independently verified. Report an error.
- MRR
- $100K–$150K portfolio (self-reported)
- ARR
- $1.2M–$1.8M portfolio (self-reported range)
- Users
- Multi-product
- Founded
- 2021
- Team
- 3 (plus AI agents)
- Stack
- JavaScript, TypeScript, React, Next.js, Node.js, Stripe, AI agents
How did the portfolio start?
Simon Høiberg launched FeedHive in February 2021 as a social media management and scheduling tool with AI features. He had already tried and failed with an earlier startup that made no money. Before founding, he freelanced (including work that once reached high rates) and had background as a LEGO engineer and React Native educator via Galaxies.dev.
Early on he made a classic mistake. At roughly $300 MRR he hired a five-person team because “that’s what you do.” Three months later he was about $50,000 poorer and had to let everyone go. He has repeated this story publicly many times as the most expensive lesson: you do not need a big team to build a real business. Systems and automation matter more than headcount.
FeedHive grew steadily without hockey-stick hype. Public posts and third-party recaps put it around $10K MRR in year one and near $20K MRR by 2022, with server costs still under $400/month at that level. He stopped sharing exact product-level revenue figures after a point, preferring to talk about the broader portfolio.
Expanding the stack
Instead of pouring everything into one product, Høiberg added complementary tools. LinkDrip (link shortening and engagement) launched in October 2022. He acquired TinyKiwi in January 2023. Aidbase, an AI-powered customer support tool, spun out in November 2023 from internal tooling he had built to handle FeedHive support load. FounderStack arrived in November 2024 as a bundle/lifetime-deal packaging of the products. SignupGate followed in September 2025.
He has described the income streams in public posts: recurring subscriptions from FeedHive, LinkDrip and Aidbase; one-time sales from FounderStack and TinyKiwi; YouTube (ads and brand deals); X; consulting; and investments. In an October 2025 post he said the combination was generating +$100K/month. A later 2026 post put the range at $100K–$150K/month. He has stated that the SaaS side alone makes six figures monthly and that he sells zero courses. These are self-reported figures. Third-party sites have published lower estimates for individual products (for example GetLatka’s ~$228K ARR estimate for FeedHive as of late 2024), which are not confirmed by the founder.
Lean operations and AI
The team size trajectory is documented in his own posts: larger groups in 2023–2024, down to 4 people in 2025 and 3 in 2026 (himself as CEO, a UI/UX designer/developer, and a tech support agent). He runs the business from Switzerland and has been setting up a Singapore entity. He describes the Swiss operation as almost 100% agentic and the new Singapore one as AI-native from day one.
AI agents handle large parts of product work, content, support, and even fully autonomous blog publishing experiments. He has moved infrastructure off big US cloud providers toward self-hosted setups and has been fine-tuning open models for specific tasks (support, content, etc.) so they can replace paid API calls for routine work. Recent commentary focuses on “agentic revenue” — income driven end-to-end by AI agents with minimal human involvement — which he says generated +$25K in a recent month as a proof of concept.
Distribution has included his large X audience (around 165K+ followers), YouTube channel (hundreds of thousands of subscribers), influencer-style content, and paid video retargeting. He is bullish on video and influencer channels alongside building in public.
Lessons that still feel true
Hiring early and large is expensive. The $50K lesson forced a permanent preference for tiny teams, automation-first workflows, and systems over people. Portfolio construction reduces risk: multiple smaller revenue systems mean one product’s struggle does not end the business.
Self-reported transparency has limits. He has publicly said that after a certain MRR level founders should stop constant revenue-sharing contests and show other signals of health instead. Exact current product-level MRR is not publicly detailed by him; the portfolio ranges above are his own statements.
AI is shifting the operating model. What used to require more headcount can now be handled by agents and fine-tuned models, enabling location freedom and lower fixed costs. The trade-off is new complexity in evaluating and training those systems, plus the usual platform and model risks.
The work remains product- and systems-focused rather than pure hype. He continues shipping updates, experimenting with pricing (including single-tier experiments on FeedHive to handle AI usage without usage-based billing), and documenting the journey without claiming perfect linear growth.
Caveats
Revenue figures cited here come from the founder’s own X posts and are labeled as self-reported. Individual product ARR estimates from third-party sites are estimates only. Exact current user counts, precise MRR per product, and full cost structure are not publicly detailed. Team size and automation claims are based on his posts. The business is bootstrapped and multi-product; outcomes are not guaranteed for others copying the approach.
Current shape of the work
Day-to-day the focus stays on product reliability, customer support automation, and content systems rather than constant new launches. Pricing experiments continue — for example testing a higher single-tier plan on FeedHive so AI features stay useful without introducing usage-based metering. The portfolio approach lets attention and trust from one product (or the personal brand on X and YouTube) support the others.
Infrastructure decisions reflect the same lean bias: leaving expensive cloud providers, self-hosting where practical, and training task-specific models on open weights. These choices reduce variable costs and increase control, but they require ongoing technical effort. The Singapore base is intended to support more AI-native operations from the start.
Support and enterprise-style conversations still exist even with agents. Documentation, clear boundaries, and selective human involvement keep the load manageable. The public narrative emphasizes process and systems over viral growth stories.
This is not a single-product rocket ship. It is a deliberately diversified, low-headcount operation that compounds through repeated shipping, content, and automation. The numbers that are public are the ones the founder chose to share; the rest stays internal.