Benchmark · MRR
A good MRR, in percentiles: 435 products measured
Every article explains what MRR means. Almost none says what a real number looks like. Here are 435 products that publish MRR: the median is $396, a quarter earn $50 or less, and the top quarter starts at $6,732. The distribution is not a curve — it is two cliffs.
Revenue as of 2026-10-30 · sample: 851 products that publish revenue · 435 report a non-zero MRR
Products reporting MRR
435
of 851 in the library
Median MRR
$396
top quarter starts at $6,732
Earning under $100/mo
33.6%
146 products
The short answer
- $396 is the median, not a target. Half of the products that bother to publish MRR sit below it.
- $1,000 MRR already puts a product in the top 59% of this sample — 179 of 435 clear it.
- $10,000 MRR is the top fifth (21.4%). Above that, the sample thins out fast: only 93 products report five figures.
Percentiles, not averages
Averages are useless here because a handful of products carry millions. Percentiles answer the question people actually have: "where am I?"
| Percentile | MRR | What it means |
|---|---|---|
| P10 | $15 | one in ten products earns this or less |
| P25 | $50 | the bottom quarter |
| P50 | $396 | the middle — half are below |
| P75 | $6,732 | the top quarter starts here |
| P90 | $28,326 | one in ten products earns this or more |
| P95 | $54,135 | the small club |
| P99 | $238,620 | near the ceiling of the sample |
Computed on the 435 products that report a non-zero MRR. A product with no reported MRR is excluded, not counted as zero.
The middle is empty
This is the part that surprises people. The distribution is not a bell curve you can aim at; it is a pile at the bottom and a long shoulder at the top, with very little in between.
| MRR band | Products | Share |
|---|---|---|
| $1–99 | 146 | 33.6% |
| $100–499 | 77 | 17.7% |
| $500–999 | 30 | 6.9% |
| $1,000–4,999 | 52 | 11.9% |
| $5,000–9,999 | 34 | 7.8% |
| $10,000–49,999 | 69 | 15.9% |
| $50,000+ | 24 | 5.5% |
Same denominator (435). 33.6% earn under $100; only 18.9% land between $500 and $5,000; 21.4% clear $10,000.
Why this matters for pricing
By category
Category medians are pulled around by a handful of large products, so read the counts next to them: a category with four products above $1k is not a market, it is an anecdote.
| Category | Products | Reporting MRR | Median | Top 10% | ≥ $1k |
|---|---|---|---|---|---|
| AI | 205 | 112 | $208 | $10,376 | 32 |
| SaaS | 49 | 23 | $70 | $18,331 | 5 |
| Marketing | 45 | 22 | $649 | $51,476 | 10 |
| Dev Tools | 41 | 14 | $93 | $1,048 | 2 |
| Mobile Apps | 37 | 30 | $67 | $10,738 | 6 |
| Content Creation | 25 | 15 | $1,667 | $9,990 | 8 |
| Education | 25 | 16 | $84 | $17,154 | 3 |
| Productivity | 24 | 13 | $22 | $106 | 1 |
Only products with an assigned category. 235 products in the library have no category and are excluded from this table — they are not missing from the percentile tables above.
By team size
Team size is declared for only 93 of 851 products, so this is a sketch, not a law. What it does show: the one-person median is a fraction of the two-to-five median.
| Team | Products | Reporting MRR | Median MRR | Top 10% |
|---|---|---|---|---|
| 1 person | 76 | 37 | $64 | $1,242 |
| 2–5 people | 17 | 13 | $362 | $16,424 |
Products with no declared team size are excluded from this table.
MRR vs ARR vs NRR
| Metric | What it is | What it hides |
|---|---|---|
| MRR | Monthly recurring revenue — subscriptions normalised to one month. | Fees, refunds, churn, and whether the subscriptions renew. |
| ARR | MRR × 12. A reporting convenience, not a separate measurement. | Everything MRR hides, plus seasonality inside the year. |
| NRR | Revenue from the same customer cohort vs a previous period, including expansion and churn. | Nothing we can compute: it needs cohort data we do not collect. |
We publish MRR only. Anything labelled ARR on this site would be arithmetic on the same numbers.
Products across the spread
One product from each band, taken from the same sample. Every row links to that product’s own fact page — the numbers there carry their own timestamp.
| Product | Category | MRR | Revenue · 30d |
|---|---|---|---|
| AssistantCoach | Health & Fitness | $98 | $98 |
| Intentsly | Sales | $97 | — |
| act101 | Dev Tools | $97 | $38 |
| AudioX | AI | $986 | $925 |
| AlgoFuse.ai | AI | $980 | $9,822 |
| Arc Raiders Map | Utilities | $961 | $1,092 |
| Adle | AI | $9,852 | $10,086 |
| AI MONEY GROUP | AI | $9,697 | $2,108 |
| SiteJourney | AI | $9,359 | $9,365 |
| Post Bridge | Unclassified | $49,600 | $34,250 |
| VirtualJK | Unclassified | $48,500 | $70,167 |
| Vid.AI | AI | $47,000 | $68,000 |
| Stan | Unclassified | $3,570,000 | — |
| Chatbase | Unclassified | $864,000 | $645,833 |
Questions people actually search
The questions below are the autocomplete phrasings people use, not the ones we would have written.
What is a good MRR for a startup?
In our sample of 435 products that publish MRR, the median is $396. A product at $6,732 is already in the top quarter, and $28,326 puts it in the top decile. "Good" depends on stage, but the distribution is unforgiving: 33.6% report less than $100 a month.
What is MRR in SaaS?
Monthly recurring revenue: the normalised value of subscriptions billed every month, excluding one-off fees. We use the figure the product itself reports through its payment provider, not a model estimate.
What does MRR stand for?
Monthly recurring revenue. It is a revenue measure, not a profit measure — it ignores payment processing fees, refunds, hosting and support costs.
MRR vs ARR vs NRR?
ARR is MRR × 12. NRR (net revenue retention) tracks the same cohort of customers across periods and includes expansion and churn. We publish MRR only; ARR would be arithmetic on the same numbers, and NRR cannot be derived from our fields, so we do not claim it.
Is $10k MRR good?
Yes: 93 of the 435 products (21.4%) report $10k or more. $10k is roughly the top fifth of this sample, and it is the point where a product usually stops being a side project.
Average MRR per customer?
We do not publish revenue per customer: active subscription counts exist for part of the sample, but mixing declared MRR with partly-declared subscriber counts would produce a number that looks precise and is not.
Method: 851 products in our library that publish revenue through a payment provider, crawled 2026-10-30. 435 of them report a non-zero MRR; medians and percentiles are computed on that subset only. Missing values are never treated as zero. This is the population that chooses to publish revenue — not the whole market.