How to evaluate any news app — including this one
Aggregation and journalism are different jobs. Problems begin when a product blurs which one it is doing.
Updated
Different jobs in the information stack
Journalists gather facts, interview sources, verify documents, and publish under a masthead that accepts legal and editorial risk for being wrong. Aggregators and briefing products organize attention across many mastheads — deciding what to surface, in what order, with what context.
Both are legitimate and they are not substitutes. A reader served only by aggregation eventually has nothing to aggregate, since the reporting has to be funded by someone. A reader served only by individual newsrooms drowns in tabs. The failure mode worth watching for is a product doing the second job while presenting itself as the first.
The attribution test
Can you tell who reported a story before you commit to it? A product built to serve readers puts the publisher's name where you can see it, because knowing the source is a large part of deciding whether to spend attention on the item.
Products built to capture attention tend to obscure this. The publisher name appears in small grey type, or after the click, or not at all. Watch also for language that quietly claims the work: a summary presented as the product's own reporting, or a byline for staff who did not do the reporting. The question is not whether attribution exists somewhere in a footer; it is whether it is available at the moment you decide.
The hosting test
Does the product send you to the publisher, or reproduce the article on its own pages? This is the sharpest dividing line, because it determines who is compensated for the reporting.
There is a spectrum. A headline and a one-sentence teaser with a prominent outbound link is ordinary indexing. A full article body reproduced on the aggregator's domain, wrapped in the aggregator's advertising, captures the value of work someone else paid for. Between them sits the summary so complete that clicking through becomes pointless — which is the practice most disputes between publishers and platforms are actually about.
The test as a reader: after using the product, do you know where to go for the full account, and is going there easy? If the design makes leaving feel like a failure state, the product is optimizing against the reporting it depends on.
The original content test
Does the product make anything of its own, and is it clearly labelled as such? Original material is not required for an aggregator to be useful — good curation is real work. But a product that publishes original writing should mark it unmistakably, with an author and a date, so you can tell whose judgment you are reading.
The specific thing to be wary of is unlabelled synthesis: machine-generated summaries of other outlets' reporting, presented in the same visual register as reporting, with no indication of origin. That combination gets you the confidence of an article with none of the accountability, and it is now cheap enough to produce at scale.
The advertising test
Where does the product place ads relative to where the value was created? Ads next to a product's own original work, or next to tools it built, are straightforward. Ads on a page whose entire content is another publication's headline and teaser are monetizing someone else's reporting.
This is a live enough issue that Google's own advertising policies address it directly, prohibiting ads on screens whose content is copied from others without added value. As a reader you can apply the same test without any policy document: look at what is on the page besides the ad, and ask who made it.
Four questions, applied to us
It would be evasive to publish this framework and not run our own product through it, so:
- Attribution — publisher names and timestamps appear on every headline card before you click. We think this passes.
- Hosting — we do not reproduce publisher article bodies. Story briefs carry a headline, the source, a short feed teaser, and a link out. Earlier versions of this site did extract full article text; we removed that.
- Original content — the Insights library is written by our desk, with author and dates, and is kept visually distinct from third-party headlines.
- Advertising — Google-served ads run on our original Insights pages and on the Markets tools we built. They do not run on the aggregated headline feed or on story-brief pages, because that content is not ours.
The honest caveat: a curated headline feed is a modest contribution on top of other people's reporting, and calling it anything more would be overselling it. The value we can legitimately claim is selection, attribution, market context, and the original explainers. The reporting belongs to the publishers, and the appropriate place to read it is their site.
What to do with the framework
Apply it to whatever surfaces news to you — this site, a search feed, a social platform, a newsletter, an assistant that summarizes your morning. The four questions are quick and they separate products that route attention honestly from products whose economics depend on intercepting it.
And where a product passes, follow the links. The single most useful habit for anyone who cares about news continuing to exist is to actually arrive at the publishers doing the reporting, and to pay some of them. Aggregation is a convenience layer. It cannot generate what it organizes.