Most brokers find out they bought resold data long after it would have been useful to know. The test is simple, it takes an afternoon, and you can run it on data you already have.

The overlap test, step by step

• Tag every lead with its source on import. If you are not already doing this, nothing else here works — start today and run the test in 60 days.
• Normalise phone numbers to digits only, and lowercase every email, before you compare. Formatting differences will hide real duplicates and you will conclude everything is fine.
• Look for the same number or email appearing under two different sources.
• Any overlap at all between two vendors who both sold you exclusivity means at least one of them is not telling you the truth.

The real cost is not the duplicated spend. It is two weeks of rep time on merchants who had already been called by nine people, and it is the point where a broker stops believing paid leads work at all. That last one is the expensive part, because it is usually wrong and it takes a year to unlearn.

What are the warning signs before I buy?

• They cannot name the traffic channel. A vendor buying from somebody else and reselling cannot tell you what the merchant was promised either — which means your rep finds out on the call.
• They also sell aged data. Ask directly. A vendor with an aged-data line has a financial reason to resell your lead eventually.
• Exclusivity has a window. "Exclusive for 30 days" means resold on day 31.
• They will not let you dial an unassigned lead live. This is the tell that matters most.

What if I only use one vendor?

Then the cross-source test is unavailable and you have to read the leads themselves. None of these signals is conclusive alone; two or three together usually are.

• Merchants who know the script. "Is this about the funding application?" on the first ring, from someone who sounds rehearsed, means they have had this conversation already today.
• Immediate rate comparison. A merchant who opens by quoting a factor rate has an offer in hand from somebody.
• Irritation disproportionate to one call. People are not usually annoyed by a first contact about something they requested ninety seconds ago.
• Clustered submit timestamps. Records that all claim to have been submitted within the same few minutes, day after day, suggest batch generation rather than live traffic.
• Contact rate far below what the vendor quoted. If they told you eighty percent reachable and you are seeing thirty, something upstream is different from the description.

What do I do if I find overlap?

Do not lead with the accusation. You want the explanation before you decide, and there are innocent versions — both vendors may buy from the same aggregator without knowing it, which is a different failure from deliberate double-selling.

• Pull the specific duplicates: phone, email, both source tags, both delivery timestamps.
• Send them to each vendor separately and ask for the generation source and timestamp of that record.
• Compare the two answers. If both claim to have generated it, at least one is reselling upstream inventory.
• Ask for replacement of the duplicates under the exclusivity term. A vendor with a real policy will do it without argument.
• If the answers are evasive or the timestamps do not reconcile, stop the order rather than negotiating. The overlap you found is unlikely to be all of it.

How do I stop it happening again?

Most of the protection is in the agreement and the intake, and both are cheap to set up once:

• Require a stated number of recipients per record, and make it one, with no expiry.
• Bar resale in aged, repackaged or derivative form explicitly. This is the clause that catches the second-life problem.
• Ask whether the vendor generates the traffic themselves. A vendor who does cannot accidentally sell you somebody else’s record.
• Tag every source at import, permanently. The test only works on data you have already been tagging, so the cost of starting today is nothing and the cost of not having started is sixty days.
• Re-run the overlap check quarterly, not once. Vendors change suppliers.