Editorial Policy

What Mora Discover is

Mora Discover is a synthesized news feed for ecommerce operators — Shopify-first, but covering DTC strategy, the creator economy, social commerce, AI for ecommerce, influencer marketing, brand building, and macro retail. We do not break original news. We pull from established outlets and community sources, cluster duplicate coverage of the same story, and publish a single careful synthesis per cluster with every source linked.

Synthesis methodology

Every story in the feed begins with a cluster of source items. The pipeline starts by polling news RSS feeds, brand posts, Reddit and Hacker News threads, YouTube uploads, and curated social posts every hour. Items that cover the same underlying story are grouped using a combination of URL canonicalization, embedding similarity, and entity overlap; clusters of two or more items are eligible for synthesis.

Synthesis runs through Gemini 3 Flash with a custom prompt that requires every paragraph to cite the specific source IDs it draws from. The model is instructed to attribute claims, mark uncertainty, and refuse to synthesize when the cluster is too thin or the sources contradict on a primary fact. Output that fails those checks is held back; it does not publish silently.

Source standards

We accept sources from three buckets:

News outlets with a documented editorial process and a contact route for corrections (e.g. Modern Retail, Retail Brew, Glossy, The Information, Marketing Brew, the Shopify partner blog).
Community sources like Reddit and Hacker News, treated as primary community sentiment but not as fact attribution.
Brand and creator posts, treated as primary statements from the brand or creator about their own work — never as analysis of someone else's.

Sources that mass-produce LLM-written content without disclosure, that have a documented track record of fabrication, or that exist primarily as paid placement vehicles are excluded.

Attribution

Every published story includes the full cluster source list — not just the items the synthesis cited, but every item that informed the cluster. Inline citation markers ([1], [2], etc.) tie specific paragraphs to the specific source they draw from. The JSON-LD on each page exposes those sources via isBasedOn, with the correct schema.org subtype for each source kind so structured-data validators and AI engines can reason about the citation chain.

Corrections

We correct errors. If a synthesized paragraph misattributes a claim, misreads a primary source, or fails to update when the underlying story develops, we revise the published story and bump the dateModified field. Material corrections are noted at the bottom of the article body with a brief note explaining what was changed and when.

Source flags and correction requests can be sent to hello@mora-marketer.com with the subject line “Discover correction”. They are read by JT directly and triaged within 48 hours on weekdays.

What we do not do

We do not publish sponsored content in the Discover feed; sponsored placements would appear in a clearly-labeled separate surface, and as of this writing none exist. We do not publish news we cannot attribute; if a cluster collapses to a single source, the synthesis carries an explicit single-source note. We do not run AI-generated boilerplate intros on topic pages — those are hand-written.

Editor

Mora Discover is edited by JT, the founder of Mora. Every byline on the feed is JT's because JT signs off on the editorial pipeline — not on every individual paragraph in real time, but on the prompt, the source list, and the daily review pass that catches synthesis errors before they propagate.

For the technical disclosure on what AI does and does not do in the pipeline, see AI Disclosure. For methodology specifics including source list and refresh cadence, see About Discover.