The board is up 31%. The median company is flat.

Method & honesty note — read this before the numbers

Source = TrustMRR's live /api/v1/startups, pulled 2026-07-06. Cohort = the top-1,000 startups by rank; of those, 800 report both a real subscription MRR (>0) and a growthMRR30d reading — that 800 is the analyzable universe for every figure here. growthMRR30d is a ratio (last-month growth), so the absolute dollar delta is derived as mrr × g / (1 + g) — i.e. we back out what the MRR was a month ago and take the difference. Three honesty flags, because they change how you should read everything below:

• Percentage growth is noise below ~$1k of base. A startup going from $10 to $135,899 reads as +1,298,336% — technically true, analytically garbage. The dataset's mean growth is +5,659% and is meaningless for exactly this reason; it's dragged by a few brand-new listings coming off ~$0.
• Anonymous listings are a data artifact, not momentum. Several rows are literally named "Anonymous startup" with absurd percentages — these are new-to-the-board listings, not real month-over-month rockets. We keep them in the dollar aggregate (the dollars are real) but flag them by name so you don't read the percentage as a signal.
• Dollar-weighted is the only honest lens. Because the low base poisons percentages, everything here leans on absolute $/mo added, plus breadth (what fraction grew vs shrank). The median and the dollar total tell you more together than either the mean or the headline alone.

Why "up 31%" and "median flat" are both true — the barbell

A single aggregate number hides its own distribution. When you split the 800 startups by whether they grew, held, or shrank, the board is barbell-shaped: mass at both ends, a hollow middle. A few big companies compound hard on the right; a nearly-equal mass bleeds on the left; the typical company in between went nowhere.

This is the core teaching point of the whole read. The aggregate went up 31% because dollars concentrate; the median went nowhere because bodies don't. "The market is growing" is a claim about weighted dollars; "the typical founder had a flat month" is a claim about counts. They don't contradict — they're measuring different things, and a healthy analyst keeps both on the table. If you only quote the +31%, you've quietly adopted the winners' point of view and erased the 38% who lost ground.

Concentration is the real story

Take the gross MRR added this month and ask who added it. The answer is brutally top-heavy: the ten largest gainers account for 52% of all the MRR added across 800 companies. This is a power-law index — the same shape as a venture portfolio, where a couple of names carry the fund and the rest wash out.

Why this matters for reading the board: an average lifted by ten names is not evidence of a broad opportunity. If you're a new entrant, the relevant question isn't "is the category growing?" (the index says yes) — it's "am I in the compounding tail or the flat middle?" (the breadth says most aren't). Concentration also makes the aggregate fragile: Stan alone is +$1.07M of the +$3.04M. Remove the single biggest mover and the "market" grows meaningfully less. One company's good month is a third of the story.

Who actually moved — the gainers, by name

Dollar-weighted, largest absolute add first. The interesting split isn't gainer-vs-loser — it's incumbent compounding (Stan) vs genuine breakout off a small base (Kibu, Postiz) vs new-listing artifact (the Anonymous rows).

• Stan — +$1.07M/mo (+42.8%, now $3.57M MRR). The single biggest absolute mover on the whole board, and the most instructive: this is an incumbent still compounding at scale, not a rocket off zero. A creator-monetization platform adding a million a month on a $2.5M base is a different, rarer, more durable thing than a $10→$100k pop. One company = a third of the cohort's entire net add.
• Kibu — +$230k to $234k MRR (+5,283%, Education US). The standout genuine breakout — a real rocket, not an artifact (it's a named company with a category). This is what a low base looks like when it's real: the percentage is huge, but so is the dollar landing.
• Postiz — +$155k to $162k MRR (+2,090%, Social Media HK). Notable twice over: it's the one non-US name near the top, and it's an open-source social-scheduler — relevant to oll.am because Postiz is already on our own self-host deploy shortlist. OSS-as-distribution is visibly converting to MRR here.
• Upscale System / AEO Engine / Supergrow / RankAI / Adspirer — a tight cluster of AI & marketing tooling landing at $55–90k MRR off +2,400–3,700% growth. This is the "AI tooling actually converting" pocket — small bases, but real dollars, mostly US, mostly SEO/marketing-adjacent.

And who gave it back — the losers, by name

Half the board shrank; here's where the biggest dollars left. Note the pattern: the losses cluster in e-commerce and analytics/attribution — categories sensitive to ad-spend cycles and one-time-purchase churn.

• Brand On Demand — −$544k/mo (−74%, still $190k MRR, E-commerce US). The biggest absolute loser on the board — a print-on-demand / e-commerce play that more than halved in a single month. A mirror image of Stan: a big base moving hard, this time down.
• Cometly (−$197k, −50%, Analytics) and GoTall (−$105k, −65%, Health). The attribution/analytics and quantified-self cohorts giving back hard — exactly the categories that swell during an ad-spend boom and deflate when budgets tighten.
• The Anonymous rows cut both ways. Just as they inflate the top, several appear here with absurd negative percentages — again, listing churn, not real collapse. Treat their percentages as noise; the named companies above are the real signal.

Where the money was added — and shed — by category

Net $/mo added, by category. Two honest asymmetries jump out: Content Creation "wins" almost entirely because of Stan (one company), while Social / SaaS / Dev Tools are healthy by breadth (85–88% of their members growing) even though their dollar totals are smaller. Velocity and revenue are different axes.

Where money was shed: only three categories are net-negative — E-commerce (−$468k), Analytics (−$171k), and Security (−$27k). E-commerce and analytics are the month's net losers, and (see the loser table) it's concentrated in a few big drops rather than broad bleed. Everything else is net-positive in dollars, even where breadth is mixed.

Geography — the US is the board

Of the +$3.04M net added, the US contributed $1.75M — more than half, from one country. The rest is a long, thin tail, and a couple of the "country" lines are really one company.

One tie-back to oll.am — then straight back to the guardrail