The article documents a consistent pattern across nursing homes, ambulances, ER staffing, and housing: PE firms acquire with 70-90% debt, load it onto the target, extract dividends, and exit in 5-7 years. Cites the 2021 NBER finding of roughly 1-in-12 excess deaths among PE-owned nursing home short-stay residents and documented cases of PE-owned ambulance operators refusing to respond to 911 calls when unit economics turn negative.
Argues that the rollup model that gutted ambulances and nursing homes is now being run against developer infrastructure — Thoma Bravo (Sophos, SolarWinds, Anaplan, Coupa, Darktrace), Vista Equity (Jamf, Ping, Pluralsight, Datto), Permira (Zendesk, Genesys), and Silver Lake (Splunk-adjacent) have rolled up large swaths of the B2B SaaS stack. The model doesn't require the acquired business to improve, only to survive to exit, which means tool degradation, capex cuts, and reduced engineering invest
Frames the core insight as structural rather than moral: PE returns come from financial engineering (debt, dividends, multiple expansion at exit), not operational improvement. This explains why the same destructive pattern recurs across wildly different industries — from dialysis to mobile home parks to back-office software — because the incentive is identical regardless of what the company actually does.
A long-form investigation making the rounds on Hacker News (489 points and climbing) lays out, in uncomfortable detail, how private equity has quietly become the landlord of American essential services. The piece documents PE ownership across emergency dispatch, ambulance fleets, ER staffing, dialysis chains, nursing homes, mobile home parks, single-family rentals, and increasingly the back-office software that runs all of them.
The numbers in the source are blunt. PE-owned nursing homes show measurably higher mortality rates than non-PE peers — a 2021 NBER working paper pegged the excess deaths at roughly 1 in 12 short-stay residents. PE-owned ER staffing firms (Envision, TeamHealth) were the dominant force behind the surprise-billing crisis that Congress finally legislated against in 2022. PE-owned ambulance operators have, in multiple documented cases, simply stopped responding to 911 calls in counties where the unit economics turned negative.
The pattern is consistent: acquire with 70-90% debt, load the target with that debt, extract dividends, cut staff and capex, sell to the next fund (or the public markets) inside 5-7 years. The model doesn't require the business to improve — it requires the business to survive long enough for the exit.
If you're reading this on top10.dev, you might wonder why a story about ambulances and nursing homes belongs on a developer publication. The answer is that the same playbook is now being run against the boring software infrastructure your stack depends on, and most engineering orgs are not pricing the risk.
Look at the recent transaction history. Thoma Bravo alone has rolled up Sophos, SolarWinds (pre-breach, notably), Anaplan, Coupa, Darktrace, and a dozen smaller security and devops vendors. Vista Equity owns or has owned Jamf, Ping Identity, Pluralsight, Datto, Cvent. Permira took out Zendesk and Genesys. Silver Lake controls Splunk-adjacent businesses. The B2B SaaS market is in the middle of the largest PE roll-up in its short history, and the targets are not random — they are sticky, mission-critical, high-switching-cost tools that your platform team can't rip out in a quarter.
The community reaction on the HN thread is telling. The top-voted comments are not from policy wonks; they're from engineers describing what happened after their vendor got taken private. A recurring sequence: contract auto-renewal terms get worse, the European support team gets laid off, the public roadmap goes dark, the changelog slows from weekly to quarterly, integrations with adjacent tools quietly break and don't get fixed, and 18 months in the renewal quote arrives with a 40-300% increase. One commenter, a former Pluralsight engineer, described post-Vista layoffs that cut the platform team to a skeleton crew while marketing budget grew. Another, working at a hospital, described how their PE-owned EHR vendor responded to a critical security CVE: a four-month delay and a price hike.
The deeper structural issue is time horizon mismatch. Infrastructure software — the kind that runs payroll, identity, monitoring, logging, CI, source control — needs decade-scale investment to stay reliable. PE funds operate on a 5-7 year clock, and every dollar that doesn't accelerate the exit is a dollar that gets cut. That math is fine for a consumer app you can swap out in a sprint. It is catastrophic for the systems your business depends on to exist.
There is a counter-argument worth steelmanning: some PE acquisitions genuinely do clean up bloated cost structures, refocus drifting roadmaps, and produce better products. Qualtrics under Silver Lake arguably shipped more in two years than it did in the four preceding. Some of the post-Microsoft-rejection rebuilds at acquired companies have been competent. But the base rate is what matters, and the base rate — measured in support deterioration, security incidents, and unilateral price increases — is bad.
First, audit your vendor list for PE ownership. If a tool you depend on has been acquired by a firm with 'Capital,' 'Partners,' 'Equity,' or 'Holdings' in the name in the last 36 months, treat it as a known-degraded dependency. Pull the contract. Find the auto-renewal clause. Calendar the notification window.
Second, price the migration. For every PE-owned critical vendor, you should have an honest 90-day estimate of what it would cost — in engineer-weeks and dollars — to move off. Not a vague 'we could switch'; a real plan with a named alternative and a migration runbook. The vendors that have you over a barrel know exactly how much it costs you to leave, and they will price the renewal accordingly. Closing that information asymmetry is the only leverage you have.
Third, favor open-source equivalents where the trade-off is acceptable. Not as ideology — as risk management. Postgres cannot be acquired and gutted. Prometheus cannot have its European support team laid off. Self-hosted Sentry, self-hosted GitLab, self-hosted Vault — these have real operational costs, but they also have no exit clock. For the layer of your stack that is genuinely critical, the calculus has shifted.
Fourth, if you're at a company doing vendor selection in 2026, read the cap table. A growth-stage company with founder control and patient capital is a fundamentally different risk than the same product owned by a fund in year 4 of a 7-year hold.
The PE roll-up of developer tooling is not slowing down — interest rates have stopped rising, dry powder is at record highs, and the public SaaS multiples remain compressed enough that take-privates pencil out. Expect more acquisitions across observability, CI/CD, security, identity, and data tooling through 2026 and 2027. The defensive move for engineering organizations is to stop treating vendor selection as a procurement problem and start treating it as a supply-chain risk problem — the same way you'd think about a sole-source hardware supplier or a single-region cloud dependency. Your monitoring vendor going dark is not an abstract scenario anymore. It's a Tuesday.
Top 10 dev stories every morning at 8am UTC. AI-curated. Retro terminal HTML email.