What is prospecting? The definition, and where the hours actually go

In sales, prospecting means deciding who's worth contacting before you contact them. You start with a list of names, check each one against what a good customer looks like, keep the few that hold up, and let the rest go. Most of the skill is in the letting go.

The word comes from mining, and the old meaning is a better guide than most sales definitions. A prospector moves a lot of gravel to find a little gold, and nobody ever judged one by how much gravel they shifted.

Prospecting vs lead generation: what's the difference?

Lead generation produces names: ads, content, a badge scan at a conference, a list bought from a data vendor. At the end of it you have rows in a spreadsheet.

Outreach contacts names, by email, phone, direct message or a comment under someone's post.

Prospecting is the judgement in between, and it's usually the first thing to get squeezed because it doesn't feel like progress. Buying a list can feel like buying prospecting, but what you've really bought is raw material, whichever lead generation tool it came from. If you skip the filtering, the recipients do it for you, in about a second each, and their verdict lands on your sending domain rather than on one message.

One way to tell which job you're doing: lead generation is judged by volume, outreach by how many messages went out, and prospecting by what you decided not to send. Prospecting matters most in outbound, where you choose who to contact. Inbound vs outbound sales covers when each one is the better start.

The four steps of prospecting

1. Decide who qualifies. Write a filter, not a description. "Mid-sized healthcare companies" is a description. "Physiotherapy clinics with two to nine practitioners and no online booking on their site" is a filter, because two people applying it to the same list will get the same answer. That's an ideal customer profile question, and everything else on this page depends on it.

2. Build the raw list. Almost any source will do. Lists got cheap, which is exactly why they stopped being an advantage: your competitors can buy the same rows.

3. Check each one is still true. Open the site and confirm that what the list says is true today. This is where the time goes, and it's the real work.

4. Write down the reason. One sentence per company that survives, with a link to where you found it. If you can't write that sentence, the company didn't survive, whatever the filter said.

Step three matters more than it looks, because sources go stale faster than people expect. Across the conversations Revtive's agents flagged as worth replying to between July and September 2026 (843 of them, across six workspaces), only 39% came with a date at all. Of the ones that did, the median post was 51 days old by the time it was found, and about a quarter were more than eight months old. That's why we cap old threads at low priority in the product, and it's why a list that looked right last quarter needs checking again before anyone writes to it.

That's a rough estimate rather than a measurement, but the shape holds: sending is the smallest part of the hour, and it's the part most outreach tools automate.

Sales prospecting techniques, compared

Each of these has its own cost and yield, and they don't line up the way you'd hope. The last row causes the most arguments. IP-based visitor tools can usually name an organisation at best, and for anyone browsing on home broadband or mobile data they often can't get past the internet provider. They can work for large companies on corporate networks, so check the match rate on your own traffic before you build a process around it.

How long does prospecting take?

If each account takes ten to twenty minutes of proper reading, fifteen accounts is a morning. That's a useful number to plan around, and it barely changes whether you do the research yourself or review something a tool drafted, because reviewing is the same reading.

So be sceptical of anyone promising sixty well-researched accounts a day from one person. Either the research is thin or the "well" is doing a lot of work. The only real lever is reducing how many accounts need the full treatment, which comes down to the filter, not the research.

A worked example

Here's an illustration of how a list shrinks. It's made up, but the proportions are realistic. Say you sell booking software to independent physiotherapy clinics, and you've bought a list of 400.

• Filter on the profile. Two to nine practitioners. The list's practitioner count is wrong often enough that you check it against each clinic's own team page. 400 becomes 212, and around forty of the drops were bad data rather than bad fit.
• Filter on the problem. Does the site already have online booking? If it does, you'd be asking them to switch, which is a different message and usually a harder sale. 212 becomes 61.
• Filter on timing. Something that makes this month the right month: a new practitioner announced, a second location, reviews complaining about phone hold times, a post asking for software recommendations. 61 becomes 19.
• Filter on the sentence. For each of the 19, write one line naming the specific thing, with a link. Running these checks cheapest-first is most of the method. Six give you a sentence you'd be happy to have quoted back to you in a reply. The other thirteen could have been written about any clinic.

Six messages from 400 rows can feel alarming the first time. It's the right ratio, though. The other 394 weren't wasted; they were the cost of finding the six.

One of those six sentences might be: "You added a second physio in August and your booking page still sends everything to the front desk phone." It's checkable, it's about them, and it's the reason the email exists at all. That's the difference between a message that gets a reply and one that gets deleted.

What prospecting won't fix

It won't tell you where to look. For a service business, the quickest prospects often aren't on any list at all. They're the people posting "can anyone recommend a..." this week, and finding them is a routine of its own.

It won't sell a product nobody wants. If six carefully researched messages get six polite refusals, the problem is further upstream, and more filtering won't find a buyer who doesn't exist.

It won't scale in a straight line. Going from six good messages a day to sixty means ten times the research or a lower standard of evidence, and in practice it's nearly always the second.

And it won't make you known. Cold outreach reaches people who've never heard of you, and they're suspicious for exactly that reason. It helps a lot if you're also answering questions where your buyers already ask them, so the cold message lands with someone who half recognises the name.

If a tool does the checking for you, hold it to the same standard you'd hold a person to: it should show the source and the date for every claim. The evidence standard doesn't change because a machine did the reading. That's how Revtive's prospecting agent is built, with a source link on every signal, because a confidence score without a link is just a guess with a number attached.