The evidence test: how to tell a real lead from a plausible one
Most lead lists are confident guesses. One question separates a prospect worth contacting from a row in a spreadsheet: what did you actually see?
A lead list is easy to produce and hard to trust. Filters return rows, rows look like progress, and nobody finds out the rows were wrong until a month of outreach has produced nothing.
There's one question that sorts a real lead from a plausible one, and it's not about the company. It's about you: what did you actually see? Not what you inferred, and not what a filter implied, but what you saw, somewhere you could link to right now.
Where do lead lists go wrong?
Most lead qualification is a chain of reasonable-sounding guesses that add up to nonsense.
The company is in construction. Construction firms deal with permits. Firms that deal with permits need help with filings. So this firm needs help with filings.
Every step sounds fine, and the conclusion is still a guess. By the time it reaches a spreadsheet the chain is invisible, so nobody remembers it was a guess at all.
The evidence test breaks the chain. For every lead, you should be able to point at something specific: a job posting, a thread, a page on their site, a filing, a public complaint. Something with a link and a date. Without that, you don't have a lead. You have a company you've heard of.
Three kinds of evidence worth separating
Not all evidence is equal, and treating it all the same recreates the problem in a subtler form.
The strongest is when someone at the company has said it out loud: a post asking for help, a job description naming the responsibility, a public complaint about the tool they use now. Nothing is left to guess.
Next is structural evidence, where something about the company makes the problem close to certain even though nobody has said so. They run twelve locations with no regional manager, or they're hiring their first salesperson. The problem follows from the shape of the business.
Weakest is circumstantial evidence: they fit a pattern that tends to go with buying, like the right industry, size or region. That's where most lead databases stop, and on its own it tells you very little.
Sort your list by those three and work from the top. A day spent on ten leads where someone said it out loud is usually worth more than a week on two hundred circumstantial ones, and you'll learn faster whether your positioning is right, because the people you're talking to actually have the problem.
Does the date of a buying signal matter?
As much as the signal itself. A hiring post from last week and one from fourteen months ago aren't the same signal, even though a filter treats them the same. The role gets filled, the project ships, the person who complained switches tools and forgets about it.
We learned how easy this is to miss in our own product. In August 2026 we found a workspace holding a Reddit thread that was more than five years old, rated high-impact, because the agent had judged how well it matched the product and not how old it was. We now cap anything older than eight months at low priority. For hiring posts and similar signals we'd be stricter still: after about ninety days, treat them as circumstantial however strong they looked when they were fresh.
Write the reason down
The habit that makes this stick is simple. For every lead you keep, write one sentence naming what you saw and why it matters. It's the same sentence an ideal customer profile is built from. If the sentence is hard to write, the lead is weak. If it turns out to be "they're in the right industry", you've found a row pretending to be a prospect.
That sentence also becomes your first line of outreach, which is a useful side effect. Qualifying a lead and writing to it turn out to be the same work.
What it costs you
Fewer leads. A lot fewer.
That can feel like going backwards, especially if you've been judging your week by how many rows you added. But forty leads you can each justify will tell you more, and usually convert better, than four hundred you can't. It's also the standard any AI outbound tool should be held to, and why a source link matters more than a confidence score when you're comparing AI SDR tools. Applied one step earlier, it's what separates a buying signal from an intent score.
If you apply the test and almost nobody passes, the method hasn't failed. The market is telling you something about your positioning that the four hundred rows were hiding.