What this is

Reputation Engine is a brand-monitoring tool used by a marketing agency on behalf of small businesses. It reads public conversations — forum posts, news articles and public feeds — looking for ones that mention a client or describe a problem that client solves, and shows the agency the ones worth a reply.

What is collected

Public posts and articles

The title, the text, the link, the source it came from, the public username of whoever posted it where one is shown, and the date. Only from places that are already public and readable without logging in.

The agency's own notes about those posts

A relevance score, a category, and any note a member of the agency types while reviewing an item. These are the agency's opinions about a post, not information collected from anyone.

What is not collected

• No accounts for readers. Nobody signs up, and there is no login for the people whose posts are read.
• No tracking pixels and no analytics. There is no analytics service of any kind in this product.
• No private messages, no private groups, nothing behind a login. If it cannot be read logged out, it is not read.
• No purchased data, no mailing lists, no contact scraping.

Cookies

Opening a shared client link sets one cookie, named repengine_share. It holds only that link's own identifier, so the page remembers which client's results you were looking at. It lasts 90 days. It is not an advertising cookie and it is not shared with anyone.

The weekly and monthly reports set no cookies at all, and neither does the sign-in page. Agency staff who sign in receive a session cookie, which is the ordinary mechanism for staying signed in.

Requests to other companies

Being exact about this matters more than sounding clean.

The weekly and monthly reports are entirely self-contained. They load nothing from anywhere, which is why they survive being emailed or printed.

Separately, finding the posts in the first place means searching for them. Search terms — a client's business name, the keywords it is monitored for, and its town or region — are sent to the public search services the tool reads from, which today are Google News, Bing News and a public archive of Reddit. Those services see the search text and the request, as they would from anyone typing the same words into a search box. Nothing about the client's customers, and nothing collected earlier, is sent with them.

Where the information is kept

On the agency's own computer hardware, in four places: the main database file, a small separate file recording how much the local models were used, nightly backup copies of the database, and any report files that have been saved out. All four sit on machines the agency owns. None of it is held in a cloud service, and none of it is sold, rented, or shared with any third party.

The software that reads and scores the posts runs on that same hardware. No text is ever sent to an outside AI service. Everything is processed on machines the agency owns.

How long it is kept please read

Indefinitely, at present. This is the part such notices most often get wrong, so it is stated plainly.

The tool only ever shows the last 30 days of activity. That is a display rule, not a deletion rule: material older than 30 days stops appearing in the product but stays in the database. There is no automatic clean-up and nothing is deleted on a schedule.

Information is removed only when a person deletes it deliberately — for example when a client is removed, which deletes that client's collected posts with it, or when someone asks for a particular post to be taken out.

Reading only, never posting

This tool has no account on any platform and no password for any platform. It cannot post, reply, vote, comment or sign in anywhere, because there is nothing for it to sign in with. Where it suggests wording for a reply, a person reads that suggestion and decides whether to post it themselves, from their own account.

Asking for something to be removed

If a post of yours has been collected and you would like it taken out, write to CONTACT-EMAIL-TO-BE-SET with a link to the post, and it will be removed from the database.

Removing it here has no effect on the original: that stays wherever you posted it. Only the agency's stored copy is deleted.

Two honest caveats. Removal is done by hand by the agency rather than by an automatic process, so it takes as long as it takes someone to do it. And because the database is backed up every night, a copy of the removed material can persist in those backups for a short period after it has gone from the live system.

Where this sits legally

The agency is in Alberta, Canada, and Canadian federal and provincial privacy law applies to it. Many of the businesses it works for, and many of the people whose public posts are read, are in the United States, including California.

The precise obligations that follow — and how they treat information a person has chosen to publish in a public forum — are being reviewed by the agency's legal counsel. This notice will be updated when that review is finished.