How we work
We publish comparisons of learning apps while selling one, so the useful thing is not a promise to be fair. It is a published method you can check, and rules strict enough that breaking them stops the site from shipping.
Learn anything, one lesson a day.
The short version
Most editorial policies are a statement of intent. These are assertions in the code that builds this site. If an article names a product without linking its official source, quotes a competitor's price, omits what a product is bad at, or passes 180 days without a source review, the build fails and nothing publishes.
That is a deliberately awkward thing to commit to, which is exactly why it is worth publishing. Below is what each rule is, followed by what we do not and cannot do.
The standards
When an article names another company's product, the description comes from that company's current official pages, and the link sits in the article. The build fails if a product is described without its source, or if that source is missing from the article's source list.
Prices move by region, platform, billing term and promotion, so a figure published here would be wrong within months. Articles describe how a product charges and link its checkout instead. The build rejects a currency figure in a competitor's pricing description.
Each entry in a roundup says what the product is bad at as well as what it is good for. A roundup where everything is excellent is an advertisement. The build fails if either is missing.
Learning Cadence publishes these articles and makes Cadence. That disclosure appears directly under the byline, not in a footnote at the end, and the build fails without it.
Several of these articles recommend Khan Academy, Coursera, Duolingo or Blinkist over our own product for specific jobs. An evaluation must name at least one alternative that is not Cadence, and must state who should not buy the product under review.
Where an article cites evidence, it names the authors, the journal and the year, and links somewhere you can read it. Widely repeated numbers we cannot trace to a primary study are called out as untraceable instead of repeated. Charts show only the figures a paper actually reports.
Each article states when its sources were last checked. The build fails once that date is more than 180 days old, so a stale article breaks the site rather than sitting there quietly going out of date.
Every roundup, evaluation and research article states its method and its limits, including which products we did not buy and which behaviour we could not check.
Our limits
Corrections
Email support@learningcadence.app with the article and the specific claim. We correct factual errors in the article itself rather than quietly deleting the page, and we update the review date when the sources are rechecked.
If a company believes we have described its product inaccurately, send the official page that says otherwise and we will correct the article and its source list.
Questions
The Learning Cadence editorial team, which is the same team that builds the product. Articles are published under the company rather than an individual because the standards on this page, not a personal reputation, are what the work is accountable to.
Partly you should not, and that is why this page exists. Every article states our commercial interest directly under the byline, links the official source for every product it names, and states what our own product is bad at. Where a different app is the better answer, the article says so. Check the sources rather than taking the conclusion.
Articles are organised by the job a reader is trying to do rather than by a ranking, because the products in this category mostly do not compete directly. Each article states its selection method and its limits in a section near the end.
Each article carries the date its sources were last checked. The site build fails once any article passes 180 days without a review, so out-of-date content blocks releases instead of accumulating.
Email support@learningcadence.app with the article and the specific claim. Corrections to a factual error are made in the article itself, and the review date is updated when we recheck the sources.
Last reviewed