SShortSingh.
Back to feed

Study of 542 failed mental health startups reveals seven recurring failure patterns

0
·12 views

A new analysis examined 542 digital mental health companies that exited the market between 2000 and 2026, coding each on up to 18 fields including business model, funding, clinical evidence, and reason for closure. The research was prompted by high-profile failures like Pear Therapeutics, which collapsed in 2023 despite FDA clearance because no health plan agreed to reimburse its product, having once been valued at $1.6 billion. Data was drawn from sources including Crunchbase, CB Insights, and trade press, with every company verified against at least two independent sources. Around 58% of the companies in the sample are from the United States, meaning the findings most accurately reflect the English-speaking digital health market. The study was conducted by the founders of Mentalium, a voice-based CBT diary app, who undertook the research before committing further resources to the mental health technology space.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Log in to join the discussion and vote.

Log in

Related stories

0
ProgrammingDEV Community ·

How to generate isometric and hexagon grid paper for print with accurate spacing

A developer explains how to produce isometric and hexagon graph paper for printing, working in PDF points where one inch equals 72 points. A common error is spacing isometric lines by the triangle side length rather than its height, which distorts the triangles. The fix is to anchor all three line families to a shared lattice point at the page centre and round coordinates to avoid duplicate edges. For hexagon grids, the article clarifies that a tabletop '1 inch' usually means the across-flats width, which implies a side length of about 0.577 inches. Finally, it recommends centring whole cells within the margins so the grid looks balanced.

0
ProgrammingDEV Community ·

osFoundry lets Codex CLI run as a worker on a user's Mac via ChatGPT login

A desktop agent in the osFoundry app can use Codex CLI as its model, launching codex exec on the user's Mac under their existing ChatGPT login. Work runs inside a folder the user has authorized, and Codex's reasoning, commands, and final message appear as lines in the osFoundry run. No OpenAI API key is involved, since the app strips API-key and base-URL variables from the child process so the run stays on the subscription, and cost is recorded as zero. Tools such as memory search are served through a loopback MCP bridge, and only when a git worktree isolates the run. Setup involves installing Codex CLI, logging in once, selecting the agent in the External AI settings, and choosing a folder for it to work in.

0
ProgrammingDEV Community ·

Free Servers Forget Everything: Five Myths About Ephemeral Deployments

Developers often assume free compute and model endpoints retain state, but such resources are typically shared and disposable. Apparent persistence may just reflect landing on the same machine during a lucky restart window. The article offers a stdlib-only Python probe that writes a small JSON state file and compares results across runs and restarts. Its results show that persistence can be confused with same-host luck, and that even surviving state needs write-then-read verification. It also argues that job state such as counters, cursors, and sent markers belongs in an external store rather than in process memory, and that chat endpoints are request-scoped, so context must be resent by the client.

0
ProgrammingDEV Community ·

From Cricket Bans to Binding Pockets: Building Mental Models for Viral Glycoproteins in VR

From Cricket Bans to Binding Pockets: Building Mental Models for Viral Glycoproteins in VR The notification hit my phone at 11:47 PM on a Tuesday. EU bans house cricket protein in food products — EFSA cites insufficient safety data on allergen cross-reactivity. I stared at the screen, thumb hovering. Three years of grant proposals, late-night sequencing runs, and that one spectacular failure with the cryo-EM grid — all circling back to a regulatory footnote about insects I'd never even held. My lab works on viral glycoproteins.