How to find an old article you read months ago when you only remember the gist

You may remember a Spanish town offering empty houses to remote workers, an explanation for aching knees at night, or only the Tuesday in October when you read it. The title and publication are gone, so three guessed Google queries return nothing useful.

Knowledge workers spend about 2.5 hours a day looking for information they already have, according to IDC figures cited by Cottrill Research. Recovering one old article is the personal version of that problem.

Why the article becomes hard to find

Reading is quick; deliberate saving is slower. Without reinforcement, the Ebbinghaus forgetting curve suggests that roughly 70% of new information can be gone after twenty-four hours. Titles and sources often disappear first, while the gist, a vivid sentence, or an image remains.

That leaves a meaning-shaped memory facing tools built around exact strings. Browser history, conventional search, and many note apps cannot reliably interpret “that piece about the Spanish village.”

Sources can disappear too: pages are rewritten, paywalled, or removed. Mozilla announced in May 2025 that Pocket would close on October 8, 2025 after 18 years, as reported by TechCrunch. A recovery workflow therefore depends on whatever trace remains indexable.

Where to look based on how it was saved

• Only read, never saved: Start with Google My Activity and a history-by-date extension. Local Chrome history is commonly limited to about 90 days; My Activity can retain a longer server-side record when you were signed in with Web & App Activity enabled.
• Browser bookmark: Sort the bookmark manager by date added. Expect title and URL search, not full-content recall.
• Pocket: Look for an earlier HTML or CSV export, then consider importing it into another read-later tool. Pocket closed on October 8, 2025, and its export window has passed for most users.
• Raindrop or Instapaper: Search their indexed page text when you remember an exact phrase.
• Notion or Evernote: Search the app, then check any tags you applied at capture time. Long archives and drifting tag habits make this less reliable.
• A content-indexing memory app: Describe the topic in ordinary language. It can only retrieve material you actually saved into it.

If the only clue is a visual, see how to find a photo by describing it. If the article was never saved, work through the broader guide to finding an article you read but did not save. For a buried Hacker News favorite, use the workflow for finding an old Hacker News story by memory.

If you never bookmarked it

Try these fallbacks in order:

1. A browser extension such as History by Date, which lets you scan a day visually.
2. Google My Activity, filtered by date and product.
3. Email, newsletter archives, Trash, and All Mail.
4. A focused natural-language query in a search engine that summarizes results.

If the article never entered any of those channels, it may be functionally lost. Saving in the moment takes seconds; reconstructing the trail later can take far longer.

One Hacker News user described years of bookmarks that were rarely useful because items could not be found again or their sites had vanished (comment, August 30, 2025). The problem is not simply forgetting to save—it is saving without a dependable retrieval surface.

The fragmentation is real: Asana's Anatomy of Work Index reported an average of nine apps per day and 25 daily app switches. The article may have crossed several of them before disappearing.

What recall-by-meaning changes

Conventional archives assume that your later query will reuse words from the page or title. Months later, you remember what the article came to mean to you instead.

A recall-by-meaning layer indexes the saved page, text in images, and voice-note transcripts, then matches your description to the content. You describe the item as you would to a friend.

dEssence follows that model. Save through the Chrome extension, Telegram bot, or web app at dessence.ai, then ask without maintaining folders or tags.

Honest limitations

A nearly empty account cannot recall much, because quality depends on what you saved. dEssence has no native iOS or Android app, team or shared-list feature, fully offline storage, or end-to-end encryption. Capture is available through Chrome, Telegram, and the web app.

Free is $0 forever with no card, up to 20 saves a day, and 10 watched things checked daily. Pro is $10 a month with up to 1,000 saves a day, 100 watched things checked hourly, and advanced search.