In this article
What was dropped, and how was it judged?
On 19 October 2026 a software services firm we work with ran one lead search. 226 searches found 2,090 posts, 455 of them already seen, and Mach judged the 1,635 new ones. It dropped 1,565 and kept 70. The full run is in one real lead search.
Every post answered the same fixed questions: is the author a person, buying or selling, our kind of work, able to pay, how urgent, in the right region. Unsure answers stay unknown, and unknowns count neither for nor against a company. Each answer keeps the post's own words as evidence, and locked rules turn the answers into a grade.
Why were 1,565 posts dropped?
The table lists every recorded reason with its count. Named reasons cover 1,564 of the 1,565 drops. One recorded no reason, and we say so instead of hiding it.
| Reason | Posts | Share of drops |
|---|---|---|
| Seller, not buyer | 497 | 31.8% |
| Talk without an ask | 339 | 21.7% |
| Real ask, wrong work | 188 | 12.0% |
| Copy of another post | 179 | 11.4% |
| Job post | 160 | 10.2% |
| Recruiter | 90 | 5.8% |
| Out of region | 78 | 5.0% |
| Freelancer gig | 21 | 1.3% |
| No need shown | 7 | under 1% |
| Public body | 3 | under 1% |
| Already filled | 2 | under 1% |
| No reason recorded | 1 | under 1% |
The first two account for 836 of the drops, 53%. A list that is too short is usually short for one of these two reasons rather than for lack of searching.
What do the three biggest reasons look like?
Seller, not buyer: 497 posts. The author sells the same work the firm sells, so the words match the firm's offer exactly. Reading that as demand is the classic mistake. A day after the second run, a stricter seller rule set 9 of its 87 keepers aside; its story is in why we hide leads that might be selling the same thing.
Talk without an ask: 339 posts. News and opinion that mention the topic but ask no one for help. A judging pass a day earlier had marked 234 of 276 posts as freelancer requests, and most were news or opinions. The lead-search post tells that story and the rule it left.
Real ask, wrong work: 188 posts. A real request, but not the kind of work the firm sells. These are the hardest to see in a count, because the ask is genuine.
Why does dropping most of the finds matter?
Time is the scarce thing. Salesforce's 2022 survey of 7,775 sales professionals found that reps spend 28% of the week actually selling. [1] The Bridge Group's 2025 survey of SDR teams found a median of 112 activities a day for 4.1 quality conversations, about one in 27. [2] Every lead that should have been dropped costs minutes from a week that is already mostly not selling.
A dated, live event matters too. In a 2025 study of nearly 4,000 buyers, 6sense found that first contact came when the buyer was already 61% through the buying process. [3] The figure reaches us second-hand through a CustomerThink summary.
Why were most keepers not ready to contact?
Of the 70 kept, 21 were ready to contact and 49 were worth a look. On the next day's run, 52 of 87 kept leads were worth a look, and the checks they missed most were being strong enough (31) and having a named person to reach (29). That second run is in the second day of a daily lead search.
Copy the drop reasons, not our numbers
Write the drop reasons before the first search, and make the system record one for every dropped lead. A drop with no reason cannot be checked, and the next change to your rules cannot be tested against it.
Review the keepers rather than every find. A person can read 70 graded leads in an afternoon. Your rates will differ with your market and your bar. The shape should not: most finds dropped for named reasons, and every keeper carrying a dated event.
Common questions
What is the most common reason a lead is dropped?
In one real run on 19 October 2026, it was that the author sold the same work as the firm: 497 of 1,565 drops, about one in three. Talk without an ask came next, with 339.
Why drop so many of the posts a search finds?
Most posts that match a topic are not from a buyer with a live need. In that run, 70 of 1,635 judged posts were kept, about 4%, and 21 were ready to contact. Time is limited: reps spend 28% of the week selling [1], so a long list of poor leads costs the time that should go to good ones.
Does a dropped post stay dropped?
Yes, unless something changes. Mach stores each post once per firm, by its address, with its grade and its reason. On the second day, 409 of the 415 finds that were posts from the first day had been dropped the first time. They stayed dropped, each with its reason, and cost nothing to judge again.
What does 'talk without an ask' mean?
A post that mentions the topic but does not ask anyone for help: news, opinion or commentary. In that run it was 339 of the 1,565 drops.
