Why accounts get rejected: what 802 rejects show

About one in four. Across four production research runs, S1 agents decided on 1,051 candidate companies and kept 249, an acceptance rate of 24%. The rate held steady from run to run: 22%, 26%, 24% and 24%. Put the other way, for every account kept, about three other companies were checked and turned down.

A candidate here is a company that surfaced in searches built from the profile, and that was not already on the client’s list or on an exclusion list. These are not random companies: each one surfaced because something about it matched the profile, which is why the reasons they failed are worth reading.

• Key takeaway: fit alone is a weak filter. Most rejected companies were turned down for timing, not for fit.
• Key takeaway: size and ownership catch most of the fit failures, and both need a source and a year to check.
• Key takeaway: a reject with a reason is useful. All 802 rejects here keep one, and 88% of those reasons cite dated evidence.

The table groups all 802 rejects by the rule each reason names. Timing dominates: 538 companies, 67.1%, had no qualifying dated event inside its window. Size outside the profile’s band came second with 115, and ownership third with 78.

Two details matter for reading it. First, industry mismatches are rare, at 2.7%, because the searches were already built from the profile’s industries. The fit failures that remain are the ones a filter cannot see: real size and current ownership. Second, the reasons carry evidence. 709 of the 802 reasons (88%) include a date, and 78 (10%) include a link. The median reason is 19 words long.

No dated event inside its window (67.1%)

The company may fit, but nothing says now is the time. A buying event, such as a new executive, a new owner or a new plant, makes a purchase more likely for a limited time; after that, the window closes. Outreach to a good-fit company with no live event tends to read as generic, because there is nothing specific to say. These accounts are not dead. Keep them with their reason and check again when a new event appears.

Size outside the band (14.3%)

The company is too small or too large for what you sell. Size is easy to get wrong, because ranges in a record are often estimates with no year. Ask for a size statement with its source and year, such as revenue from a 2025 filing or headcount from a state business filing. For private companies, decide in advance what happens when the figure cannot be found.

Ownership (9.7%)

The company is a subsidiary that buys through its parent, has already been acquired, has a sale pending, or breaks another ownership rule in the profile. A record can say “independent” long after a deal closes. A quick search for “acquired” or “to be acquired” with the company name catches most of these before anyone writes to them.

Industry, duplicates, geography and committee (the rest)

Industry mismatches mean a company that looked right sells something else, for example a distributor listed as a manufacturer. Duplicates are the same company under another website. Geography rejects a headquarters outside the target region. The committee bar rejects a company where too few of the people who decide could be confirmed from a recent source.

Take a sample of your current list and judge it the same way: accept, or reject with one reason. Then count the reasons. The counts tell you which part of the list is broken.

1. Pick 50 companies from your list at random, not the ones you already like.
2. For each, record what it sells, size with source and year, ownership today and headquarters.
3. Look for a dated event that your profile treats as a buying signal, and check that it is still inside its window.
4. Accept the company, or reject it with the first rule it fails, written the same way every time.
5. Count the reasons. If timing leads, the market fits but the moment does not: keep those accounts and set a date to re-check them.
6. If size or ownership leads, the source of the list is the problem. Change where companies come from, not the outreach.
7. If one rule rejects almost everything, check whether it reflects how you really win, and change it only with a dated note.

This is early data, and it is narrow. It comes from four runs on two profiles, and three of the runs used the same client’s ICP, which targets US mid-market and enterprise companies. Another profile, with different signals and windows, would produce a different mix of reasons.

• The figures describe what the research recorded. They are not a blind or human audit of each reject.
• Short rule codes at the start of each reason are new: only the latest run used them, for 84% of its rejects. Earlier runs wrote plain sentences naming the rule.
• The “other” row is mostly an error from early runs that the current research brief no longer allows.
• Timing rejects depend on the windows the profile sets. Longer windows would turn some of them into accepts.

In SalesOne, S1 agents check every candidate against the structured ICP: what the company sells, size with its source and year, ownership, headquarters and, when the request requires one, a dated event inside its window. A rejected company is saved with the rule it failed and the evidence, and SalesOne refuses a reject without a reason. That is why all 802 rejects in this data keep one.

When reject reasons show a pattern in a target group, SalesOne suggests the ICP change, and one click sends it to S1 agents, which publish a new ICP version for the next run. S1 agents do the homework. Your team builds the relationship.