2026-05-16
May 16, 2026
Mentors talk to you. Sponsors talk about you. In a market where US tech job listings sit 36% below their 2020 baseline and middle-management is being compressed, sponsorship is the only career lever that consistently compounds — and most senior TPMs and tech leaders are still optimizing for the wrong one.
This is the part of the calendar where the senior people I know start asking the same two questions. "Should I be looking?" And, more quietly, "Why isn't the work I'm doing translating?" The brag doc is current. The peer feedback is glowing. The H1 review went well. And yet — promotion comes up empty, or the lateral that was supposed to materialize hasn't, or the role at the next company sits in a recruiter pipeline for nine weeks before quietly closing.
The instinct is to do more of the same thing harder. Update the brag doc. Schedule another mentor coffee. Add one more bullet to the impact memo.
The diagnosis is almost always the same. The mentor count is high. The sponsor count is zero. And in a 2026 hiring market where overall US tech job listings sit roughly 36% below their February 2020 baseline and global tech hiring rates hover around 29%, sponsorship is the only lever that consistently moves a Staff+ career.
The distinction has been formalized for over a decade and most senior tech leaders still get it wrong. Carla Harris, former vice chairman at Morgan Stanley, named it cleanly in her 2018 TED talk: mentors talk to you, sponsors talk about you. Sylvia Ann Hewlett's research at the Center for Talent Innovation, published as The Real Benefit of Finding a Sponsor in HBR (2011) and later as a book (HBR Press, 2019), put the same idea on a quantitative footing: sponsored employees were materially more likely to ask for stretch assignments and to be promoted than peers with mentors alone.
Lara Hogan's Resilient Management (A Book Apart, 2019) operationalized the distinction for engineering orgs as a three-mode framework: mentors give advice, coaches ask questions to help you find your own answer, and sponsors put their reputation behind you in rooms you are not in. The same human can play all three roles — but the work they do in each role is different, the asks you make of them are different, and the dynamics that make one relationship work make the others fail.
Why senior tech leaders accumulate mentors and zero sponsors. Mentorship is offered. Someone asks if you want to grab a coffee. You say yes. Six months later you have a recurring 30-minute slot. Sponsorship is engineered. Nobody offers to sponsor you. Sponsors have a finite supply of political and social capital, and they spend it where the return is highest. If you are not visible to a sponsor, not in their problem space, not solving a problem they care about, and not making them look prescient when they back you — they will not back you, and the gap is not their failure to mentor. It is your failure to put legible work in front of them.
The 2026 market makes the asymmetry brutal. When companies are flat-hiring at Staff+ and the only sourcing pattern that beats the cold-application odds is a direct internal endorsement, mentors don't move you. The senior manager who took you to coffee three times last year doesn't have a vote in the next role. The Director two levels above your skip-level does. The principal engineer who runs the architecture review does. The VP of Engineering at the company you actually want to be at, whom you have not yet met, does — and the path to her vote is through someone she already trusts. Mentors are useful for sense-making, for navigating tricky conversations, for the hundred judgment calls per month a Staff+ role surfaces. They are not the mechanism by which your scope expands or your next role opens.
The engineering job to be done. Stop counting mentors. Start engineering five sponsor relationships, deliberately, with named candidates and a 12-month investment horizon. Hogan calls the eventual configuration a Manager Voltron — five people who together can advocate for you across the rooms you need someone in: your direct manager, your manager's manager, a peer-org leader, a domain expert at staff or principal level, and at least one external connector with cross-company reach. The Voltron is the most defensible career artifact a senior tech leader can build, and unlike titles, it travels with you across roles, companies, and economic cycles.
Three moves to make this quarter. First, list five named individuals — by name, today — who would be net new sponsors if the relationship were activated. They must be people whose advocacy materially changes outcomes, which excludes most peers and most mentors. Second, rewrite the brag doc for a sponsor audience. Sponsors don't need a list of accomplishments; they need a one-page brief they can read in 90 seconds and quote in a room 90 days later. Tagline, three load-bearing wins with quantified outcomes, the next archetype you are growing into, and the specific ask. Third, manufacture two pieces of work in the next 90 days that need a sponsor's airtime — a memo that surfaces a cross-team decision, a working-backwards doc for an initiative that touches their org, a public talk or post that lands inside their reading list. Sponsors back people whose work makes the sponsor look prescient. Make it easy.
The trap that kills most Staff+ candidates. Tanya Reilly named it in 2019 and the dynamic has not weakened: glue work — the unglamorous, coordination-heavy, fundamentally invisible labor that holds teams together. It is real, valuable, and almost completely unsponsorable. A career composed of glue work produces a brag doc full of "facilitated" and "unblocked" and "coordinated" — verbs that no sponsor can quote in a promotion committee because they describe support, not bets. The senior TPMs and tech leaders who get sponsored at Staff+ are doing legible work on top of the glue work — the migration they led, the platform they founded, the policy they wrote, the team they spun up. The glue work is the price of admission. The legible work is the sponsor ammunition.
Try this week. Open a blank doc. Title it Sponsor Brief — {Your Name}, May 2026. Write five lines: (1) the archetype you are growing into; (2) three load-bearing wins from the last 12 months, each with a number; (3) the one program you want to be associated with for the next 12 months; (4) the specific stretch role or scope you want to be in the running for; (5) the ask — exactly what you would want a sponsor to say about you in a room you are not in. Then identify the five people who could plausibly say that line if asked. That is your starting Voltron.
What it is. A three-mode framework for the developmental relationships a senior tech leader needs. Mentors give specific advice based on their experience — "here's what I did when I hit this". Coaches ask questions that help you find your own answer — "what do you think is going on?". Sponsors spend their political and social capital on your behalf in rooms you are not in — "she should run that program". The same individual can occupy more than one mode, but the work is different in each, the asks are different, and conflating them is the most common failure mode senior people have in their own career development.
When to use it. During the quarterly career reflection. Before any deliberate career move — promotion push, internal lateral, external job search, scope expansion, or archetype shift. Whenever you feel like you are "doing the work" and not getting the return: the triad usually surfaces a missing sponsor relationship that the work, by itself, will not generate.
How to run it:
When NOT to use it. Don't use this framework as a label for existing relationships and then change nothing. The triad is a diagnostic, and the value is in the rebalancing. Don't use it during an active crisis either — Hogan's own caveat is that the work of building a Voltron is slow, deliberate, and pre-emptive. If you need a sponsor today, the time to engineer one was two quarters ago; the second-best time is now, but the relationship won't pay off for two more quarters.
Worked example. A staff engineer six years into a hyperscaler has three mentors, two coaches, and zero sponsors. Her brag doc is excellent. Her promotion has not landed in two cycles. She maps the triad: her manager is a mentor (talks to her, never about her), her skip-level is unknown to her, her peer-org leaders are friendly but disengaged, the principal engineer she most respects has never seen her work. Over six months she shifts: she presents the platform reliability story to her skip-level in a memo he can quote, writes an architecture decision record the principal engineer cites in his own thinking, and proposes a cross-team initiative whose Driver role her manager publicly hands her. Her next promotion cycle is unanimous. The brag doc didn't change. The rooms did.
ICONIQ R&D Compensation & Incentives — ICONIQ's R&D comp benchmarks are the cleanest reference for late-stage compensation conversations at Staff+ and director+. Late-stage cash comp runs a median 29% above early-stage; San Francisco, New York, and Seattle carry a 20–40% premium over the national median for leadership roles. Walk into the next negotiation having priced your role against this, not against levels.fyi alone.
Tech Job Market 2026 — Pin — The bifurcation is sharper than the headline: ML engineer openings up 59% over 2020, general SWE postings down 49%, and AI staff engineers pulling an 18.7% premium over non-AI peers — a gap that widens with seniority. Read your own role through both lenses before deciding whether the market is bad or just selective.
Glue work considered harmful — Sean Goedecke — A useful counterweight to the "lean into glue work" reading of Tanya Reilly's original talk. Goedecke's argument: senior engineers who become known for glue work get kept in coordination roles permanently. Pair with Reilly's original to triangulate where the line sits for your own next promotion.
"You need a sponsor in the room when those critical decisions are being made."
— Carla Harris, How to find the person who can help you get ahead at work (TED, 2018)
Don't miss what's next. Subscribe to Critical Path: