Why hasn't everyone switched to a personal data pod yet? The cold-start problem, honestly

The critique, kept in our own archive and aimed at us

An outside review of an early Solidus plan put it without decoration: the value proposition needs users, operators and content, and you begin with none of them. Its summary of the usual outcome:

Projects launch with great tech, no users, die quietly.

We keep that document. It also asked why this should succeed when better-funded attempts struggled, and that question has not been answered by anything we have shipped.

It named specific companies as failures. We are not repeating those judgements: we did not verify them, they are somebody else's business, and a competitor's difficulty has never been an argument for us.

And one line from it we refuse even though it is about somebody else

The critique said Solid has tiny adoption after years.

That is not accurate and we are not going to launder it through a page of ours. There are real deployments, a government programme, a public data utility, an enterprise server with versioned releases. What is thin is consumer adoption, which is a different sentence and the true one.

Getting this right matters more for us than for them, because a page that overstated Solid's failure would be arguing that the model does not work, while we run on it.

Why people actually have not switched

One: the incumbent is free and works. Storage inside an application costs the user nothing visible and requires no decision. "Where your data lives" is not a felt problem on a Tuesday.

Two: the costs land on the user and the benefits are diffuse. Setting up a pod is work today, for a benefit that arrives later and quietly. That is the worst shape a product can have.

Three: the bottleneck is the application side, not the storage side. A pod is only useful when applications read from it. Every application that supports pods had to be written or rewritten to do so, and the incentive to do that work is weakest exactly when there are fewest pods.

Four: nothing breaks if you ignore it. Technologies that spread quickly usually fix something painful. This one prevents a harm most people have already absorbed.

And the reason specific to us

We make the friction worse, not better.

Generic identifier tooling does not resolve ours: an integrator has to be taught our method specifically. Our interface exposes two of the four permission modes. There is no bulk export, and no account deletion.

So on the specific question "why has nobody switched to ours", the honest answer is not the industry's cold-start problem.