HOW WE VERIFY

How we keep FendLand accurate

FendLand tells property owners who is contacting them, where their data is exposed, and how to get it removed. Those answers are only worth acting on if they are current and correct, so here is exactly how we check our own work. Where something has a limit, we say so plainly rather than imply more than we do.

We re-check every opt-out link on a weekly sweep

The removal catalog only helps if the opt-out links still work, and data brokers move, rename, and retire their pages constantly. So every week we sweep the active brokers that carry a public opt-out URL and fetch each page to confirm it is still live. When a link is confirmed, that broker is stamped with the date it was last verified. When a broker moves its form, we update the link and note what changed before re-stamping it.

That date is not hidden in a database. Each broker’s removal page shows it in plain language, for example “FendLand last confirmed this opt-out link works in” the relevant month. You can see it on any page in the removal directory. If a page has not been re-verified recently, its date says so.

A person reviews every guide before it publishes

Some of our guides and articles are drafted with the help of an AI model. None of them go live on that basis alone. Every drafted article starts as a draft that is not indexed by search engines, is kept out of the guides hub, and is not listed in our sitemap. It becomes public only after a person reviews it and approves it. Until that approval, the draft stays a draft.

Review means checking that every product claim matches what FendLand actually does, that the advice is genuinely useful and includes steps you can take without us, and that nothing is stated as legal advice. An article that does not clear that bar is rejected rather than published.

The scanner’s classifier is tested against real messages

The message scanner decides whether a text reads as a wholesaler, an investor, a realtor, or a scam. Its always-on classifier runs against a labeled set of real and canonical messages as part of our automated test suite, so a change to the scoring rules cannot quietly make it less accurate without a test catching it. The test reports where the classifier confuses one category for another, which is what we work to reduce.

We are honest about what that number is and is not. It measures the deterministic rules that run on every scan, not the extra AI check we reserve for genuinely unclear property messages, and it is a lab measurement rather than a promise about your specific text. On iPhone, the on-device junk filter scores the message content locally on your phone and does not send it anywhere.

Where we are honest about limits

Data removal is ultimately up to each broker, so we file and track requests but do not guarantee removal from any list or platform. Verification tells you a link works or an article was reviewed; it is not a promise of a specific outcome. If you ever find something on the site that looks wrong or out of date, we want to fix it: email support@fendland.com and a real person will look into it.