The micro-SaaS acquisition market is real, structured, and larger than most engineers realize. SaaS M&A hit 2,698 transactions in 2025, a 28% jump from the year prior, and the supply of sub-$10K products for sale has expanded dramatically as AI compressed the time from idea to working MVP. For a technical buyer, the market is interesting precisely because sellers are often not technical — the discount is baked into the asking price.
What the multiples actually look like
Micro-SaaS (under $100K purchase price) trades on SDE multiples: seller’s discretionary earnings, meaning net profit plus owner salary and one-time expenses added back. The Acquire.com biannual report for Jan 2026 put the median SaaS profit multiple at 3.9x SDE for 2024–2025, with clear tiering by deal size:
- $100K–$250K: ~2.1x average profit multiple
- $250K–$1M: ~2.3x average
- $1M+: ~2.9x average
These are averages and mask a wide spread. A business with flat or growing cohorts, low churn, no customer concentration, and clean financials commands a premium inside each band. A business where the seller is the product — their relationships, their domain knowledge, their manual work — trades at the bottom. That’s where the technical buyer finds the gap.
Margins in profitable listings averaged 71% in 2024 (up from 67% in 2023), which means the multiple math is forgiving: a $50K SDE business at 3x is a $150K acquisition generating ~$4K/month in owner earnings before any improvement. The improvement is where the return comes from.
The operator edge and where it applies
The core thesis is that most sellers of small SaaS products are non-technical founders. They can’t patch the Rails version, can’t migrate off a deprecated API, can’t refactor the billing logic that’s leaking revenue. These problems show up as declining MRR and hit the asking price — but they’re solvable in a weekend by someone who can read a codebase.
This is the same dynamic that makes Systems Thinking useful here: the seller sees a symptom (declining revenue), can’t identify the structural cause (tech debt, broken integration, abandoned SEO), and discounts the asset. The buyer who can diagnose the feedback loop gets paid for that diagnosis.
That edge is real but bounded. It applies where the problem is technical. It doesn’t apply where the problem is distribution (who would you sell to?), market timing (the niche is dying), or customer acquisition cost (the business never had a repeatable channel). Reading a codebase doesn’t fix a broken go-to-market.
Moat taxonomy for the AI era
The filter that matters most right now: would this product still exist if every user had Claude or ChatGPT? Most content-aggregation plays, most “AI writing tools,” most SEO-automation tools fail this test. The AI replaces the product directly.
The Vendep research on vertical SaaS moats makes the core point: the durable moat is workflow ownership, not data ownership alone. Proprietary operational data compounds only when it’s embedded in how customers actually work — when leaving the product means re-entering years of operational history and retraining staff on a new system. Data in a dead product is just a CSV export.
Four moat types ordered by durability in the current environment:
Workflow integration / switching cost. A tool that sits inside a daily process (approvals, invoicing, client reporting) survives AI commoditization because switching means process change, not just software change. This is lever-5 thinking: the rules are embedded in the system, not the surface.
Vertical niche / market density. Vertical SaaS hit ~$130B market size in 2025 growing at 18–22% CAGR, roughly double horizontal SaaS growth. The reason: products purpose-built for a specific industry accumulate switching costs that compound at scale. A product with 30–40% penetration in a tight niche has a different moat than one with 2% penetration across a broad market.
Proprietary data. Becomes a real moat when it’s generated by the workflow (not just collected) and would take years to reproduce. Industry benchmarks, anonymized transaction histories, proprietary taxonomies built by customers using the product — these resist substitution. Generic scraped data does not.
Technical difficulty. Weakest standalone moat. Another engineer can build it, just hasn’t yet. Worth counting if combined with the others, or if the technical barrier is genuinely high (deep integrations, compliance requirements, hard-to-replicate ML training pipelines). Not worth counting if it’s just “this took a long time to build.”
Score these honestly. A business that’s genuinely workflow-embedded in a vertical niche with proprietary customer-generated data is a different asset than a business that’s technically complex but switching-cost-free.
Marketplace landscape
| Marketplace | Best use | What to know |
|---|---|---|
| Acquire.com | SaaS primary | Stripe-verified listings; cleanest for bootstrapped SaaS under $500K |
| Microns | Sub-$100K primary | Zero commission, built-in escrow; purpose-built for micro-SaaS |
| Empire Flippers | $100K+ calibration | ~88% of asking achieved; browse free to learn what a clean listing looks like |
| Flippa | Volume / breadth | Widest funnel under $250K; vet hard — quality varies widely |
| FE International / Quiet Light | Path B brokers | Full-service diligence support; appropriate for $200K+ first-time buyers working with an advisor |
Empire Flippers is useful even if you never buy there. Their listings show what a properly documented business looks like: P&L with add-backs explained, cohort retention graphs, traffic analytics, operator time breakdowns. Calibrating against this standard makes it easier to identify what’s missing in a Flippa listing.
Diligence: the specific failures that cost people money
Revenue verification is the most common place where acquisitions go wrong. Screenshots are not verification. FE International’s diligence framework requires read access to the payment processor — Stripe, PayPal, Braintree — and reconciliation of every MRR figure against actual bank records line by line. A discrepancy above 10% is a walk condition, not a negotiation point. The seller either has clean data or doesn’t.
Churn is the second failure point. A single aggregate churn number means nothing without a cohort breakdown. The question is whether the retention curve flattens. A curve that goes 100% → 80% → 72% → 70% → 70% describes a business with a healthy retained core. A curve that goes 100% → 80% → 60% → 40% describes a leaky bucket where revenue is propped up by new acquisition masking structural loss.
Customer concentration is structural risk, not a preference. One customer above 20–25% of revenue means a single cancellation can trigger an inability to service acquisition debt. This should affect deal structure, not just asking price — earnout provisions tied to that customer’s retention are the appropriate response.
Tech debt assessment is where the technical buyer earns their edge. Look for: dependencies on deprecated APIs or packages with no maintainer, authentication or billing logic built on custom code rather than a standard provider, absence of tests on revenue-critical paths, and any infrastructure that exists only as institutional knowledge (undocumented manual processes, servers no one has credentials to). These are the problems sellers can’t fix and buyers need to price or fix.
Deal structure
A typical safe deal structure for a first acquisition under $250K:
- Escrow always. Escrow.com or a marketplace’s built-in escrow. No exceptions. There is no trust-based substitute.
- Seller financing / earnout for 20–30% of purchase price. Common structure is 70% cash at close, 30% in performance-based provisions paid out over 12–36 months tied to ARR retention, churn rate, or EBITDA targets. This shifts post-close risk back to the seller. A seller who inflated numbers won’t agree to earnout provisions tied to those same numbers.
- Transition period. Negotiate a defined handoff window (typically 30–90 days) with specific deliverables: code access, credentials, customer relationship introductions, documentation of undocumented processes. Put it in the contract, not email.
The Clearly Acquired seller notes and earnouts overview notes that seller financing is common in deals under $50M and serves both parties: buyers preserve capital, sellers often command a higher total price. The risk is a seller who pushes for short repayment timelines or no amortization (all principal due at end). Either of those is a red flag about their confidence in the asset’s ongoing performance.
The decision structure
Path A ($20K–$50K) and Path B ($150K–$250K) are not just price bands — they’re different decisions with different information requirements. The Decisions framework applies directly here: which is the one-way door, and does your current level of certainty justify it?
Path A is a two-way door. A $25K asset that underperforms is recoverable — you lose some capital, but you gain the information return (what diligence catches, what it misses, what operating the asset teaches you about the category). Path B is closer to a one-way door for a first deal: a $200K mistake at an unproven edge is expensive and takes time to recover from.
The right progression: Path A deal to prove diligence and operating skill, then Path B when the edge is demonstrated. Reserve 70–80% of your total allocation as dry powder through the first deal.
Try it
Marketplace calibration session (2 hours, free). Browse 20 listings on Acquire.com filtered to recurring revenue SaaS, price range $25K–$100K. For each, apply the defensibility score (0–2 per category) and note whether you can verify the MRR claim from the listing alone. Track how many auto-reject on a single red flag. The goal is calibrating your filter speed — after 20 listings, the patterns that signal a quick pass become automatic.
Cohort reconstruction exercise (1–2 hours, spreadsheet). Take any SaaS business’ disclosed revenue history (many Flippa listings show monthly revenue going back 12–24 months). Build a simple cohort model: estimate new customer count per month and implied churn to match the reported numbers. Does the math require an implausibly low churn rate? That’s the kind of thing a real diligence process would catch at the source.
See also
- Decisions — the path A/B selection is a classic one-way vs two-way door; the triage step prevents costly premature commitment
- Systems Thinking — moats as reinforcing feedback loops; the workflow-embeddedness dynamic is a “success to the successful” system trap running in your favor
Sources
- Acquire.com Biannual Acquisition Multiples Report, Jan 2026 — median multiples, deal size tiers, margin trends
- FE International SaaS Due Diligence Checklist for Buyers — comprehensive diligence framework from a specialist broker
- Clearly Acquired: Seller Notes, Earnouts, and SBA — deal structure mechanics for sub-$50M acquisitions
- Vendep: “Forget the Data Moat: The Workflow Is Your Fortress in Vertical SaaS” — moat taxonomy for the AI era
- Flippa: 2025 Online Business M&A Insights — market volume, transaction trends, category breakdown
- SaaS Mag: “Vertical SaaS Is Winning” — vertical SaaS growth rates and M&A concentration data
- Acquiring Alpha: Red Flags in SaaS Due Diligence — practitioner-level red flag taxonomy