Director of Product Staffing That Starts With the Span, Not the Title
Two numbers decide this hire. How many product managers report in, and how many product surfaces they cover. We settle both on the first call, because together they tell you whether you’re filling a group PM seat, a real director seat, or a VP seat somebody priced wrong.
KORE1 places directors of product by scoping span first, the number of product managers reporting in and the number of product surfaces they own, because that pair decides which of three very different hires you’re making. Our technology desks fill in 17 days on average, and a year after a start date 92% of those hires are still in the job.
Last updated: August 29, 2026
A founder called us in March about a director of product role. Fourth month open. Two offers already declined and a third candidate who ghosted after the final round, which stung, because everyone had liked her.
We asked the two questions we always ask. How many PMs report to this person, and how many products do they cover. The answer was two PMs and one product, and he had never counted it out before. He’d been running a director search, at a director’s price, for a job that was going to bore anyone actually qualified to do it by the end of the first quarter.
He didn’t have a sourcing problem. He had a scoping problem. No recruiter had told him.
The title sits in an awkward place. It’s the first rung where the work stops being product craft and starts being other people’s product craft, and it’s also the title companies reach for when they want to sound serious about a hire they haven’t sized yet. So the same three words get printed on reqs that are 60 grand apart in real scope. Our product desk sits inside the wider IT staffing practice and runs alongside the product recruiting team, and the first thing we do on a director search is get both numbers written down while the req is still editable.
Has This Person Actually Stopped Writing the Specs?
This is the screen. Everything else on a director resume can be learned, borrowed or coached. The willingness to let a PM ship something worse than you’d have shipped yourself, and to not quietly take it back, cannot. Nobody teaches that.
First-time directors fail here more than anywhere else, and rarely because they’re bad at product. They fail because they’re good at it and the pull is enormous.
The pattern is easy to spot once you know the shape. A PM struggles with a spec. The director rewrites it on a Saturday and calls it a one-off. The next month there is another one. By month six the PM has not improved, because nothing was ever handed back, and the director has absorbed a second job nobody budgeted for. Three of the searches we ran in the last two years were reruns of roles already filled once. All three broke here.
So KORE1 recruiters ask for the opposite story. Name something a PM of yours shipped that you disagreed with, and say what happened next. Managers who have really done it answer fast, usually with a wince, and the detail is specific. The ones who never let go stay vague. A common version is a story about coaching the PM around to the right answer, which is the same behaviour in a better jacket. The 2026 guide to hiring a director of product management walks through the full interview loop we hand clients, and the rung below this one is covered on our group product manager staffing page.
Count the PMs and the Surfaces Before You Write the Req
Every director of product req arrives with a title and a salary band on it. Almost none arrive with these two numbers, which is unfortunate, because these two numbers are the job. Orange marks the range where the title is honest. Outside that range you are overpaying for a group PM or underpaying for a VP, and neither mistake shows up until the quarter is gone.
You want a group PM
One area, a couple of PMs, and the person still owns a roadmap with their own hands. Real directors get restless here. Post the job honestly and it fills faster, and for less.
The first true director seat
Five people is the point where coaching stops being a side task. Two surfaces is the point where something has to lose. Most first-time directors are made right here, and most of them wobble here too.
Portfolio, with managers underneath
When did you last hire somebody who manages managers? That is the shift. Add a headcount plan defended line by line and a revenue number nobody can personally move, and the band moves with it.
Call it what it is
Seven surfaces, plus design and product marketing on dotted lines, adds up to a product organization. Anyone capable of running one already knows what running one pays. Our VP of product staffing page handles that search.
Directors Get Hired by Product and Undone by Engineering
Nobody screens for this and it decides year one. A director of product doesn’t report to engineering or design, and can’t tell either of them what to do, yet almost nothing on their roadmap ships unless both say yes.
So the job is political, and somebody should say that during the intake.
A candidate who thrived beside a strong engineering culture can stall out against an engineering director who treats every request as an incursion. The reverse trips people up more often, though. Strong process is not automatically the easier environment, and plenty of directors are better in a fight than in a planning meeting.
What we go looking for is one specific no. Which request did engineering refuse? We then want the next two weeks in detail, and we want to know whether those two people were still working together six months on. The useful answers are small. A director we placed last year rewrote her own proposal after the platform team told her the migration she had assumed was never going to happen. Detail at that grain is very hard to invent on the spot, which is exactly why we ask for it.
Then we ask you the same kind of question. How long have the peers been in those seats, and has anybody already failed against them? That is an awkward ninety seconds on an intake call, and it has saved more searches for us than any screen we run.
Sometimes the answers move the search somewhere else entirely. A genuinely technical roadmap with a platform org on the other side of it starts looking like the work our technical product manager staffing desk sees every week, and a mandate that turns out to be about cadence and tooling rather than strategy is usually better served by product operations staffing first.
The Same Title, Sitting in Four Very Different Companies
Span sizes the job. Who can actually do it depends on where the seat sits, and these four settings are the ones our product desk fills most often. A candidate who is excellent in one of them is regularly a poor fit for the next.
Series B to C, first product layer
The founder still has opinions, and usually a keyboard. Somebody has to build a planning process where none existed without stepping on the person who got the company this far. Diplomacy outranks pedigree.
Enterprise B2B with a real sales org
One client had eleven roadmap commitments made by account executives before the director even started. She unpicked nine of them in six weeks and kept every account. We screen for that particular nerve.
Regulated or clinical products
Validation sets the release cadence, not the team. What matters is knowing which corners genuinely cannot be cut, and a director who learns that on the job learns it expensively. Runs jointly with our regulated industry tech staffing desk.
Post-acquisition, two roadmaps
Year one is consolidation politics. Two products, two sets of loyalties, one plan. Reading about that is not the same as having survived it, so we bias hard toward people who carry the scar.
We Start on Your Org Chart, With a Pen
Print it out. Count the boxes under the seat, then count the surfaces those boxes cover. Saying both numbers out loud changes the shape of most intake calls inside fifteen minutes. Roughly a third of them end with a different title on the req than the one we started with.
The comp conversation gets easier straight after, because the band stops being a guess.
There’s useful outside evidence for why the counting matters. Pave’s compensation data, drawn from roughly 135,000 managers and executives at companies over 100 people, puts the average span for directors and senior directors at 4.6 direct reports, up from 4.3 two years earlier. So a five-PM director seat is dead average and a fourteen-PM one is nowhere near it, whatever the title says. Supply is the other half of it. Federal projections from the Occupational Outlook Handbook have technology management roles expanding well ahead of the average occupation this decade. Product is a narrow lane inside that, and the manager track is narrower still.
Once the span is settled we price it against the real market rather than a salary site average. Our 2026 director of product salary guide publishes the three tier bands and the reason six platforms report six different numbers for the same title.
Five Moves, Intake to Signed Offer
KORE1 runs the same five moves whether this is a first product leader or a replacement nobody is allowed to know about yet. What changes is how quiet step two has to be.
- 01
Span and Surface Count
Half an hour on the org chart. Boxes under the seat, surfaces those boxes cover, who the peers are, and what the last person in the job found hardest about it. Roughly a third of the reqs we take end this call with a different title.
- 02
Quiet Outreach
Directors of product are almost never on the market. We open with people our recruiters have already placed, interviewed, or turned down for something else entirely. A name somebody can vouch for beats volume outreach at this level, every time.
- 03
The Handoff Screen
One real story, walked end to end, about a PM who shipped something the candidate disagreed with. What you get back is a page of notes your VP can read in two minutes. Not a reformatted resume with a summary paragraph bolted on top.
- 04
Peer Exposure
Finalists meet the engineering and design leads early, not in the last round. Bad chemistry then shows up in week three rather than month four. Two of our searches last year changed direction entirely off one of these conversations.
- 05
Offer, Then the First Quarter
We carry the offer conversation ourselves, then call at one month, two months and three. Unglamorous work, and it is the clearest single reason our one-year retention number holds up.
Not Every Director Seat Needs a Permanent Director
You get the same recruiters, the same bench and the same handoff screen in all three. What moves is how long you need somebody for, and whether finance has actually signed the headcount off yet.
Direct Hire
Most director searches land here, because the area needs an owner rather than cover. You pay us once the person starts. We make the first-quarter calls no matter which model you end up choosing.
How direct hire works →Contract and Contract-to-Hire
Leave cover, a migration year, or a genuine need parked behind a freeze nobody can lift until Q3. It also fits the awkward case where nobody is actually certain the director seat is permanent, and that case comes up far more than companies tend to admit.
Contract Staffing →Fractional Product Leadership
Somebody who has already run the org, coming in two or three days a week. Founders at Series A and B often want this instead of the full seat, since a director’s worth of work has not turned up yet and hiring against it anyway just shortens the runway.
Fractional CPO Services →Common Questions
How many product managers should report to a director of product?
Four to six is the working range for an area director. Pave’s data puts the average across directors and senior directors at 4.6 direct reports. Below three, the seat is a group PM job with a nicer title on it.
Above eight or nine there is almost always a layer of group PMs sitting in between, which makes it a portfolio job on a different band. Fourteen reports with design and product marketing hanging off dotted lines is a VP of product wearing somebody else’s title. The people who can do that job are perfectly aware of it.
The req is approved at director level. How do we know it isn’t really a group PM job?
Ask whether the person will still own a roadmap themselves. If they will, it’s a group PM seat. A director answers for several PMs’ worth of output and doesn’t carry a backlog of their own.
The calendar is the other tell. Sketch the hire’s typical week before you post anything, and if more than roughly a quarter of it goes to spec writing, discovery calls and grooming a backlog, what you have scoped is a player-coach. Nothing wrong with hiring one. It costs less and it fills quicker once the posting says so.
Our last two directors of product came from much bigger companies and neither lasted a year. What happened?
Scale mismatch, nearly every time. At a company with 400 engineers a director inherits research, analytics, design ops and a planning cadence that already work. Move the same person into a 40-person org and none of that scaffolding exists.
Their talent was never the issue. What made them effective was orchestration, and there is nothing yet to orchestrate. So on a first-product-layer search we bias toward people who have been the entire function themselves at some point, even where that means a smaller logo on the resume than you were picturing. The smaller logo usually buys a better first year.
We have a strong internal candidate. Should we still run an outside search?
Run a benchmark search rather than a full one, and tell your internal candidate it is happening. Wharton research on external hiring found outside hires are paid about 18% more, and are 61% more likely to be let go inside two years.
That finding argues against our own invoice and it stays on the page anyway. A bad director hire costs a year, and we would rather be your recruiter in year two. What a benchmark search actually buys is calibration. Meet three or four people at the level, learn what the market is really paying, and if your internal candidate stands up next to them you can promote with evidence instead of a hunch. If they don’t stand up, the shortlist already exists. Our roundup of top tech recruiting firms covers how to size up a search partner before you commit.
Does a director of product need to be technical?
They need enough fluency to be told no by an engineer and know whether the no is real. That’s a lower bar than writing code and a much higher bar than reading a systems diagram politely.
It matters most on infrastructure, developer tools, data platforms and anything sold to other engineers, where a non-technical director tends to get managed rather than managing, usually inside a quarter. Application and workflow products are a different story. Judgment, sequencing and the ability to actually run a planning process do more work there than depth ever will. We set the bar during intake rather than defaulting to a computer science degree in the posting, because that requirement quietly filters out good people for no reason at all.
How long does a director of product search take, and what makes one run long?
Six to ten weeks is typical for director-level product, against the 17-day average across our technology desks generally. Leadership searches run longer because the pool is smaller and nobody in it is actively looking.
Three things stretch a search. An unsettled span keeps changing the shape of the shortlist. An interview loop past five stages sheds the strongest candidates somewhere around stage four, and they rarely tell you why. Then there is the comp band pulled off a salary aggregator instead of the market. Settle the span at intake and the other two tend to fix themselves.
Can you run the search while our current director is still in the seat?
Yes, and a meaningful share of the director searches we run are confidential replacements. Candidates are briefed on the company and the scope without the role being posted anywhere or the title being named in outreach.
Two practical notes from the confidential searches KORE1 has run. Keep the interview panel to three or four people and tell them explicitly it’s confidential, because the leak almost always comes from inside rather than from the candidate side. And decide early what you’ll say if the incumbent asks directly. The searches that go badly are the ones where nobody thought about that in advance.
Tell us the two numbers. We’ll tell you which seat you’re actually filling.
One call settles the span, names the peers the hire has to win over, and puts a band on the seat that finance and the market both recognise.
Talk to a Product Recruiter →