Anti-fraud in real estate: how bots eat your budget

In the ad account report a fraudulent inquiry looks like any other: a name, a phone number, a timestamp. The difference shows later — when managers spend a week calling "buyers" who never pick up. Here is how to recognise junk traffic in your reports, and which filters to put in place before you pay.

What fraud looks like in a report

Typical signs — each one a reason to look closer, several together almost a diagnosis:

— inquiries arrive in batches at the same time, often at night;
— names follow one pattern: first names only, or oddly identical structures;
— phone numbers are valid by format but never answer, or the owner "never left any inquiry";
— conversion from inquiry to conversation tends to zero, while the cost per lead is suspiciously low;
— traffic from placements that aren't in your media plan, or from geographies you never targeted;
— seconds spent on the site, yet the form is filled in perfectly and without typos.

Where junk traffic comes from

Click farms and bots — automated form-filling: how dishonest placements "deliver" their conversion plans.
Incentivised traffic — real people paid to leave an inquiry. The phone is real, the person may even confirm interest — and then vanish.
Accidental touches — not malice but bad mechanics: provocative creative, quiz forms with a prize, autofill. Formally an inquiry exists; the intent doesn't.
Recycled databases — "inquiries" from purchased lists: people who once asked about property somewhere else. Your call is cold and unwelcome to them.

Filters before payment

Technical validation: checking the number exists and its type (mobile/virtual), cutting duplicates and same-device inquiries, honeypot fields and anti-bot scripts in the form.
Behavioural filters: time to form completion, referral source, browsing depth. A four-second inquiry from the first screen is not a buyer.
A control contact: a call or message before handover to sales. Interest, budget and readiness to talk confirmed — only then does the inquiry count. This is the same buyer scoring that screens out not just fraud but plain irrelevant inquiries too.

Count buyers, not inquiries

As long as your contractor's KPI is the number of inquiries, they have no motivation to filter: any fraud improves their report. Change the unit of measurement: pay for a qualified inquiry with confirmed intent, and require irrelevant ones to be replaced. A "cheap" $10 inquiry where one in thirty becomes a conversation costs more than an "expensive" $100 one where every second person talks. How to fix that in a contract — see how an honest replacement guarantee works.

Frequently asked questions