If you work in or sell to retained executive search, you know the two numbers. A retained search takes about 123 days. About 71% of them end in a placement. They appear in pitch decks, in trade press, in vendor content and in a great many consultancy proposals.
We had the old pair on this site. It was in our own research documents and it went into an industry page before anybody checked which edition it came from. It is corrected now, and the correction is printed on the executive search page rather than quietly swapped, because a page that silently adopts the better number has learned nothing.
The finding underneath it that nobody quotes
The more useful number in the same dataset is not the headline at all. On roughly 100,000 projects, 17% of retained searches are cancelled by the client and a further 7% are terminated. Close to a quarter of retained searches do not end in a placement, and the larger part of that is the client stopping rather than the market failing.
That reframes the whole problem. The thing most likely to cost a search firm a completed engagement is not a sourcing problem, and no amount of better sourcing addresses it. It is what happens between a slate going out and a decision coming back.
The same dataset splits projects by whether the client engaged during the search, measured by client log-ins to the platform. Searches with engagement recorded came in six percentage points above the average placement rate and slightly under the average duration. Searches with none came in below average on both.
What the publisher sells, which you should weigh
The benchmark is published by a company that sells executive search software, and the engagement finding is measured by log-ins to that company's own platform. That is a real limitation and it points in an obvious direction: a vendor reporting that clients who use the software more get better outcomes is reporting something they have an interest in.
It is also correlation on a stated sample, which is more than most of this category offers, and the underlying observation is one most search partners will recognise without any data at all. Clients who stay engaged make decisions faster. The dataset is consistent with that. It does not prove the platform caused it.
The practical version
If you run searches, the number worth tracking is not the industry's average duration. It is the one nobody currently has: how many of your live searches are waiting on you, and how many are waiting on the client, right now. Almost every search platform records both states as the same word, which is why a partner cannot answer that question at a Monday meeting.
Separating those two states is a smaller job than it sounds and it changes what a pipeline review is for. That is the first system we build for search firms, and the reasoning behind it is published in full on the executive search page.
Sources
Publisher, year, sample, and what the publisher sells. The same four things every cited figure on this site carries.
Executive Search Performance Benchmark Report
- Publisher, year and sample
- Clockwork Recruiting, updated November 2024. Starting set of more than 100,000 anonymised search projects run between 2020 and 2023, filtered to retained work: non-retained projects removed, projects closed with no contract removed, projects active under three days removed, and closings beyond 400 days removed as outliers. Reports 117 days average duration and a 76% placement rate, with 17% of searches cancelled by the client and a further 7% terminated. The widely quoted 123 days and 71% are from the 2018 edition of the same report.
- What the publisher sells
- Clockwork Recruiting sells executive search software, and the client-engagement finding in the same report is measured by log-ins to its own platform. It states its sample, which most of this category does not.