Buying B2B Leads in 2026: The Honest Math (and What Works Instead)

Somebody in every B2B company eventually suggests it: "Why don't we just buy leads?" It sounds like the shortcut of the century, pay money, skip the hard part, get pipeline. This article is the honest math on what you actually get, because we run outbound for a living and still tell prospects not to buy leads. That should tell you something.

What you actually receive when you buy "leads"

The word "lead" is doing heavy lifting in these offers. What vendors sell under that name falls into three very different buckets:

1. Address lists. Company names, contacts, emails, scraped or licensed, sold as is. No interest, no context, just data of wildly varying age. Studies of B2B data decay put it around a quarter to a third per year: people change jobs, companies restructure, addresses die. A list that was accurate when compiled is measurably wrong by the time you buy it.

2. "Qualified leads" from shared campaigns. Contacts who downloaded a whitepaper or clicked an ad somewhere, sold to several buyers at once. The interest was real, briefly, generic, and it was never in you. By the time you call, two competitors already have.

3. Exclusive lead generation billed per lead. The most honest of the three, and still misaligned: the vendor gets paid per handover, not per deal, which quietly optimizes for volume at the qualification boundary.

The real cost accounting

The sticker price, somewhere between cents per address and three-digit sums per "qualified lead", is the smallest line in the true bill.

Bounces poison your sender reputation. Send to a bought list with 15 to 30 percent dead addresses and mailbox providers read it as spammer behavior. The damage lands on your domain and outlives the campaign by months. We wrote about the mechanics in our deliverability guide.

Legal exposure, especially in DACH. In German-speaking markets, outreach to bought addresses is exactly the pattern regulators and competitors' lawyers look for: no documented reason for contact, no recognizable relevance, volume over research. The rules are covered in our guide on what is legally allowed; bought lists fail almost every test in it.

Reply rates round to zero. The list is cold, shared and stale, and the message on top of it is necessarily generic, because you know nothing about these companies except that they were for sale. Sub-one-percent reply rates are the norm, and the few replies skew toward "remove me".

The hidden opportunity cost. Every week spent blasting a dead list is a week your team believes outbound "does not work", which is often the most expensive outcome of all: the channel gets written off instead of the method.

The one legitimate use of bought data

Here is the nuance the "never buy anything" crowd misses: raw company data as research input is fine. Databases, firmographic filters, technology data, trade registers, all legitimate starting material, the same way a phone book was. The line is simple: buying data to research who might have a reason to talk to you is workmanship. Buying "leads" as a substitute for creating that reason is the shortcut that is not one.

What works instead

The working alternative is not a secret, it is just work: define precisely who you are for, watch for the buying signals that create timing, build small lists yourself and verify every address before sending, then earn the reply with a first sentence that proves you did the research. Volume goes down, replies go up, the domain stays clean, and the law stays on your side. That is the whole system, and it is what we build for clients as a done-for-you engine, with every list researched fresh and owned by the client.

If you were about to spend a four-figure sum on a lead list, run the comparison first: our cost breakdown shows what a qualified meeting costs across every model. Bought leads look cheap per row and expensive per result, and the gap is not close.