Open Web vs Walled Gardens: The Numbers

Here is an uncomfortable statistic to start your morning: the average person spends hours a day inside apps they do not control, reading content they did not choose, ranked by systems they cannot inspect. The open web did not die. It just got quieter.

For researchers, this matters more than it does for casual scrollers. Your input diet shapes your output. If your information comes pre-filtered by engagement algorithms, your sense of "what's important" is being quietly outsourced to a recommendation engine.

This piece looks at the actual data on where attention flows, what closed platforms optimize for, and what you gain by rebuilding a portable, open content diet.

Where attention actually flows

The direction of travel is well documented. According to the Reuters Institute's Digital News Report, a growing share of people access news primarily through social media, search, and aggregators rather than by going directly to a publisher. Direct access — typing a URL, using a bookmark, opening a feed — has been declining for years (Reuters Institute Digital News Report 2024).

Platform data tells the same story from the other side. Pew Research Center reports that a substantial and rising percentage of adults regularly get news from social media sites, with platforms like Facebook, YouTube, and others serving as primary discovery surfaces (Pew Research Center, News Platform Fact Sheet).

The pattern is consistent: discovery has moved from destinations you choose to feeds that choose for you. This is the walled garden in practice — not a single wall, but a gravitational pull toward a small number of closed environments where the platform controls ranking, format, and what you are allowed to save or export.

For an individual, this feels convenient. For a researcher trying to maintain a defensible, reproducible view of a field, it introduces a subtle problem: you no longer own the pipeline.

What walled gardens optimize for

Closed platforms are not neutral pipes. They rank content to maximize engagement metrics — time on site, sessions, interactions — because that is what their business model rewards. That objective is not secretly evil, but it is not aligned with your goal of understanding a topic efficiently.

Three consequences follow directly from that misalignment.

Volume over relevance

Engagement systems reward content that provokes a reaction. That skews your feed toward the emotionally sticky rather than the genuinely useful. You end up collecting more and comprehending less — a gap researchers feel acutely when they realize how many "saved for later" items they never revisit.

Format lock-in

Inside a walled garden, content is shaped to the platform's format. Threads, short clips, and infinite scroll are optimized for the feed, not for careful reading or citation. Exporting a clean, portable copy is often deliberately difficult.

Opacity

You cannot inspect the ranking. When your inputs are ordered by a system you cannot audit, you cannot fully reason about your own bias. For research work, that is a methodological weakness, not just an annoyance.

The open web still works — and it's measurable

Here is the good news the discourse often skips: the open web never actually left. RSS, Atom, newsletters, and standard web pages still power an enormous amount of content. The infrastructure is boring, stable, and — crucially — portable.

Portability is the whole point. An RSS feed is a plain, machine-readable list of what a source published, in order, with no algorithm in between. A newsletter lands in an inbox you own. A saved article can be exported, annotated, and re-read on your terms. The signal-to-noise ratio tends to be higher because you set the sources, not an engagement model.

There is also a resilience argument. Open formats like RSS and OPML mean your subscriptions are not trapped inside one company's app. If a platform changes its rules — as they regularly do — a portable diet survives the change. The World Wide Web Consortium has long championed open, interoperable standards precisely because they outlast individual vendors.

For researchers, the practical payoff is reproducibility. A curated feed list is a documented, shareable methodology for how you stay current. An algorithmic timeline is not.

Building a portable content diet

Rebuilding an open content diet is less work than it sounds, and the tooling has caught up. The goal is simple: pull the sources you actually trust into one place you control, in a format you can read, listen to, annotate, and export.

Start with an RSS reader. Subscribe to the publications, blogs, and researchers whose primary work you rely on instead of hoping their posts surface in a timeline. Import your existing subscriptions with OPML so you are not starting from zero. Add newsletters to the same inbox so your reading queue lives in one portable place rather than scattered across a dozen apps.

This is exactly the problem Omphalis was built to solve on the reader side — saving articles, subscribing to feeds and newsletters, highlighting and annotating, and then listening to any of it with natural voices when you would rather not stare at a screen. It is the read, listen, comprehend half of a healthy content diet: you decide the sources, and the tool respects that.

The comprehension angle matters. A portable diet is only useful if you actually process it. Listening to your backlog while commuting or walking closes the gap between what you collect and what you consume — the core failure mode of every overflowing "read later" list.

The producer's side of the open web

Consuming on the open web is half the loop. Publishing on it is the other half — and it is where researchers can reclaim distribution they have quietly ceded to platforms.

If you write reports, run a lab blog, or maintain a newsletter, the open web lets your audience subscribe on their terms rather than depending on a platform's reach. Owning your feed means owning your relationship with readers.

Audio extends that reach. Turning a long report or literature review into narration makes it accessible to people who would never sit down and read it in full. You can convert a PDF or document into clean audio and publish a listen link alongside the text, with no platform gatekeeping who hears it. If you want to experiment with voices and pacing before committing, the EchoLive Playground is a low-friction place to start.

The principle is the same on both sides of the loop: keep your content — incoming and outgoing — in formats you control.

What the numbers actually recommend

Pull the threads together and the data points in one direction. Attention has concentrated inside closed platforms that optimize for engagement, not understanding. Direct and open-web access has declined. And the algorithmic feed, whatever its conveniences, is an un-auditable input source — a genuine liability for anyone whose job is to reason clearly about a field.

None of this requires abandoning the platforms entirely. It requires refusing to let them be your only pipeline. A portable diet — feeds you chose, sources you documented, content you can export and re-read — is a small structural change with an outsized effect on the quality of what reaches you.

The open web is still here. It is just waiting for you to point your attention back at it on purpose.

If you want to rebuild a content diet you actually control — saving articles, subscribing to feeds and newsletters, and listening to the whole backlog in natural voices — Omphalis is built for exactly that. And when you are ready to publish on the open web instead of just reading it, EchoLive turns your documents into shareable audio.