How Executive Search Works: The Retained Search Process

Last updated: August 19, 2026

By Mike Carter, Director of Partnership Success, KORE1

A retained executive search runs in installments over roughly 90 to 120 days, with the firm mapping a market, approaching passive leaders directly, and delivering three to five assessed finalists instead of a stack of applicants.

That timeline is two to three times longer than the published benchmark. SHRM’s 2025 Recruiting Executives Benchmarking report puts median time-to-fill for executive positions at 45 days. Show that number to anyone who has sat through a real CTO search and watch their face.

Both figures are honest. They count different clocks. SHRM measures requisition-open to offer-accepted across every seat an employer happens to file under “executive,” and a lot of those get filled by someone the company already knew. A mapped external search for a scarce leader has never worked on that schedule and it is not going to start now.

I take these calls before the search begins rather than during it. A CEO, usually, or a board chair who has never bought one of these and wants to know what three months of retainer actually buys. I am not neutral. Obviously. Our retained executive search practice bills for this work and I run the side of the house that signs the agreement. The version below is the one I give people on the phone anyway, quiet weeks included, because the quiet weeks are when clients call me sounding worried.

Still deciding between engagement models rather than trying to understand the mechanics? Start with our breakdown of retained versus contingency search. This piece assumes the retainer is signed and walks the calendar forward from there. For the firm-selection question, we ranked the field separately in our roundup of executive search firms.

What a Retained Executive Search Actually Is

Retained executive search is an exclusive engagement in which a company pays a search firm in scheduled installments to research a defined talent market, approach qualified leaders who are not job hunting, assess them against an agreed standard, and present a short slate. The fee is owed for the work itself, not only for a hire.

First-time buyers snag on that line. It also explains the shape of everything downstream. A firm paid for running a search correctly can afford three weeks of research nobody sees. A firm paid only on placement cannot. One economic fact, and it sets the whole calendar.

Cost is real. SHRM’s benchmarking put the 2025 median cost-per-hire for executive positions at $10,625, up 113% from $5,000 in 2017 and 21% above 2022. Nonexecutive hires ran $1,200 in the same period. In extra-large organizations the ratio between the two hit 18.3 to 1. And those figures cover all executive hiring, retained search fees included or not depending on who filled out the survey, so treat them as a floor rather than a quote.

The Seven Stages and the Clock Each One Runs On

Here is the shape of a search that goes normally. Windows overlap on purpose. Research never stops when outreach starts, and outreach never stops when your first interviews get scheduled, which is why the day counts below add up to more than 120. The stages are sequential on a slide. They are not sequential in life.

Payment usually tracks that calendar in thirds. One installment at signing, one at slate delivery or day 30, one on start date. Some firms bill a flat engagement fee plus a completion fee instead. Ask which structure you are signing, because the milestone attached to installment two is the single best indicator of whether a firm expects to be judged on process or on luck.

Stage One Is Where Most Failed Searches Are Actually Lost

The kickoff gets treated as paperwork. It is not paperwork.

A good consultant will spend four to six hours with your CEO, your board members, the two peers this person has to work with daily, and whoever held the seat last if that person is reachable and willing. They are hunting for disagreement. Say the CEO wants a Chief Data Officer who will build a platform team, while the CFO mostly wants clean revenue reporting out of a Snowflake instance that three vendors have already had their hands on. Two different hires. Two different comp bands. Deciding which C-suite seat you actually need is the work that belongs before any of this. Nobody in the room had caught it, because nobody had ever said both sentences out loud on the same day.

A firm that skips this and takes your job description at face value is doing you no favors. Job descriptions get written by committee. They describe an org chart, not a job.

Out of that session you should get a written scorecard: four to six outcomes the person owns in the first 18 months, ranked, with the ranking argued about rather than assumed. Comp band set against actual market data rather than what you paid the last person in 2022. A target list of companies. An off-limits list, which matters more than clients expect, because your largest customer probably should not learn you are recruiting their VP of Engineering.

We had a search stall out in Denver last year for exactly this reason. Head of Data Platform, funded, urgent, everyone aligned. Three weeks in it became clear the CTO wanted a builder who would write dbt models himself and the CEO wanted someone who had run a 40-person org. Same title. Two candidates who would never be the same person. We stopped the search, went back to a scorecard conversation, and restarted eleven days later against a narrower brief. That restart cost time. Discovering it at the offer stage would have cost the whole quarter.

How Search Firms Find People Who Are Not Looking

Research first. Outreach second. That order is the whole trick, and it is not much of a trick.

Before anyone gets contacted, researchers build out a named universe. That means people who already hold a comparable seat somewhere else today, at companies close enough to yours on headcount, funding stage, and how gnarly the technical estate actually is. For a VP of Engineering seat at a Series C company in Austin, that means people currently at Series C and Series D companies in Austin, plus the layer of directors at larger firms who are ready to step up, plus a handful of people who left the market for a bad reason and might come back. Titles are the worst filter available. Truly the worst. Half the useful candidates carry a title that does not match yours at all.

Then approaches go out, one at a time, from a person rather than a template. The first conversation is not a pitch. It is a diagnostic about whether the candidate’s situation has any give in it at all, and a good consultant can usually tell inside twelve minutes. Most say no. Fine. Expected, even. The ones who say “not for me, but call Priya at the health-tech company across town” are worth as much as the ones who say yes, which is why the map keeps growing after outreach starts.

Our recruiters average 15-plus years in their verticals, which mostly matters because the second call goes differently when the person on the other end recognizes that you understand their world. You cannot fake that. Candidates hang up on people who are reading from a script about a “transformational leadership opportunity.”

The Quiet Weeks, and How to Read Them

Weeks three through six are when my phone rings.

Every time.

Nothing visible is happening. No resumes are arriving. The client, who has just paid a real invoice, starts wondering whether anything is being done at all. This is normal, and it is also the stretch when a bad firm and a good firm look identical from the outside, so here is how to tell them apart without waiting until day 70.

Ask for the weekly report and look at what is in it. A real one carries a count of the mapped universe, how many have been approached, how many responded, how many are in active conversation, and how many declined with the reason attached. The reasons are the valuable part. “Comp too low against current package” appearing five times in a row is your market telling you something specific about your band, and you want to hear that in week four rather than week eleven.

A report that says “we are actively sourcing strong candidates” is not a report. It is a stall. Twice now I have told a client to fire a firm on the strength of that one sentence.

Something else worth knowing. Passive candidates move on their own schedule and it is slower than yours. A sitting VP of Engineering with a vesting cliff in March and a product launch in May is not going to have an exploratory coffee in January no matter how good your story is. Some searches sit still for ten days because the four best people in the market are all mid-cycle at once, and no amount of pressure from you changes that. Patience here is not passivity. It is arithmetic about other people’s calendars.

The Slate, and Why It Arrives All at Once

You should receive three to five finalists, delivered together, each with a written assessment. Not a trickle. Together matters, because a slate presented at once forces comparison against a standard, while candidates arriving one at a time get compared against whoever came before them, and that is how companies talk themselves into a B-plus hire in February because January’s candidate was worse.

Each write-up should cover the scorecard outcomes one by one, where the person is strong, where they are not, what the consultant probed and what came back. Comp expectations tested, not guessed. A stated reason for considering a move that has been pushed on hard enough to know whether it holds.

Read them before you meet anybody. Clients skip this constantly. Then they run a first round that rediscovers, at some length, what was already written down for them.

If the slate is thin, that is information rather than failure. Two finalists instead of four usually means the brief is too narrow, the comp band is off, or the market genuinely does not contain what you asked for. All three are fixable in week eight. None of them are fixable in week sixteen.

References, the Offer, and the Part Everyone Rushes

References at this level are not the three names the candidate hands you.

Those are the warm-up.

Those get called, sure. The useful work is backchannel, which means the consultant finds people who worked with the candidate and were not nominated by them, and asks specific questions rather than general ones. Not “was she a good leader?” Instead, how did she handle the quarter the roadmap slipped? Who on her team followed her out to the next company? What did she do the time the CFO pushed back hard? The general question gets you a warm nothing. The specific one occasionally gets you a pause on the phone that tells you more than the answer does.

Then the offer, which is where searches die more often than clients expect. The firm should have tested the number with the candidate before it is formally extended, so nobody is surprised. Counteroffers are near-certain at this level and the good consultants have already had that conversation twice, because the moment a strong executive resigns, their current employer suddenly finds the budget and the title they could not find eight months ago.

Then notice periods. Three months is common for senior leaders. Longer overseas. And a start date landing in the third month of a fiscal quarter creates its own set of problems that nobody thinks about until the offer is already signed.

Why bother with all that care? A miss is brutally expensive. Harvard Business Review has reported that a new executive has a 50% chance of leaving inside 18 months, citing an estimate that pins the cost at ten times that executive’s salary. Set that against the Bureau of Labor Statistics median chief executive wage of $206,420 in May 2024 and the arithmetic on a bad seat gets ugly quickly. Our 12-month retention across placements runs 92%. The reference stage is a real part of why.

Where Searches Actually Break

In rough order of how often we see it:

1. The scorecard was never settled. Two stakeholders wanted different people and the search ran on the average of their opinions, which describes nobody who exists.
2. Interview scheduling. A finalist waits eleven days for a second round and takes the other offer. This one sits entirely with the client, and it is the single most common way a finalist gets lost.
3. Comp band drift. The band was set on stale data, the market corrected it in week five, and nobody told the CFO until the offer stage.
4. Panel bloat. Seven interviewers, no assigned focus areas, every one of them asking about the candidate’s biggest weakness.
5. Silence between rounds. Senior candidates read a two-week gap as disinterest or as chaos. Both readings hurt.
6. The counteroffer nobody prepared for.
7. Reference theater, where the process happens but nobody asks a question sharp enough to change an outcome.

Notice how many of those sit on the client side of the table. Roughly half, in our experience, which is not the answer a search firm is supposed to give, but it is the one that helps you.

When You Should Not Run a Retained Search

Plenty of leadership roles do not need one.

Most, honestly.

A Director of Engineering in a metro with a deep bench, a comp band that is genuinely competitive, and no confidentiality requirement is a direct hire search, not a retained engagement. Same for most first-line management seats. If your last three hires at that level came from inbound applications and worked out, keep doing that.

Internal promotion deserves a real look before any of this starts. Cheaper. Faster. The person already knows where the bodies are buried. The honest test is whether you would hire that internal candidate if they walked in cold from outside today. Answer yes and you are finished here. Go tell them.

Hiring three similar leadership roles at once? A retained search per seat is usually the wrong shape. That is a project engagement.

KORE1 places across 30-plus U.S. metros and eight verticals, which means we turn down retained assignments fairly often, usually because the role does not need the machinery. Nobody at our shop gets paid for that recommendation either. It still gets made, because selling someone a $60,000 process for a $40,000 problem is how you get one engagement instead of ten years of them.

Things Clients Call Me About Mid-Search

Three weeks in and we have seen nothing. Is that normal?

Weeks three through six are the quiet stretch on nearly every retained search, because research and first-contact work produce no visible output. What should exist is a weekly report with mapped, approached, responded, and declined counts. If the report is adjectives instead of numbers, escalate.

How do search firms find people who are not looking?

Research first, outreach second. Researchers build a named list of the people already holding comparable seats at comparable companies, then consultants approach them one at a time instead of posting a role. Most say no. Their referrals often become the actual hire, which is why the list keeps growing after outreach begins.

Who actually runs our search once the contract is signed?

Ask in the pitch meeting, not after. Larger firms often sell with a partner and deliver with an associate, which is fine if disclosed and a problem if it is not. Get the name of the person doing the candidate calls and the name of the researcher building your map.

Can we keep interviewing candidates we found ourselves?

Yes, and say so upfront so the fee treatment is written into the agreement rather than argued about later. Most firms will carve out named individuals already in your pipeline at signing. What they will not accept is a candidate they surfaced being reclassified as yours in month three.

What happens if none of the finalists are right?

Short answer: the firm goes back out and you do not pay a second fee. A retained engagement covers the search, not a single slate. In practice, a rejected slate almost always means the scorecard or the comp band needs revisiting, and a good consultant will say so before restarting.

Does the firm handle the offer, or do we?

Both, in a specific order. The consultant tests the number and structure with the candidate privately, reports back what will land, and then you extend it formally. Skipping the test step is how companies send offers that get countered, stalled, or declined in front of a board.

Realistically, how much of our own time does this take?

Twenty to thirty hours of executive time across the full search, front-loaded and back-loaded. The kickoff runs four to six hours, weekly calls are 30 minutes, and interview rounds cluster around days 60 to 90. The stretch in between asks almost nothing of you.

Before You Sign Anything

Three questions settle most of it. Who does the actual work. What the weekly report contains. What triggers installment two. A firm that answers all three plainly is showing you its process, in detail, without being asked twice. A firm that answers with relationships and networks is showing you something else.

The rest is calendar discipline on your side. Decide within 48 hours of each round. Keep the panel small. And never let a finalist sit in silence while your leadership team is off traveling.

If you are budgeting a leadership hire and want to sanity-check the band before anything else happens, our salary benchmark tool will get you close, and the role-specific guides go deeper on chief data officer compensation and chief product officer pay bands. If the seat is already defined and you want to talk through whether it warrants a retained process at all, talk to our executive team. Our executive recruiting team works across technology leadership, including chief data officer and chief AI officer searches, and we will tell you when you do not need us.