Mentoring and sponsorship at the senior level

Status: drafted · Time: 35 min · Audience: staff-plus Outcome: Distinguish mentoring from sponsorship, time-box relationships, recognise mentoring as performance work, and avoid the named failure modes.

Council members mentor. They sponsor. They coach Black Belt candidates through their boss fights, advocate for under-recognised engineers in rooms those engineers are not in, and run programs that scale beyond individual relationships. This is real work, and the literature is unambiguous that it is the most under-recognised work senior engineers do.

This chapter draws on Lara Hogan’s Resilient Management and her writing at larahogan.me, Tanya Reilly’s “Being Glue” essay (the most-cited piece on the subject), Sylvia Ann Hewlett’s Forget a Mentor, Find a Sponsor (the upstream source on sponsorship), Will Larson’s writing on coaching versus mentoring at lethain.com, and Camille Fournier’s The Manager’s Path as the management-side counterpoint.

The chapter also constructs a framework for accepting versus declining mentoring requests. The literature gestures at this question without answering it, and senior contributors who do not have a framework either say yes to everyone and burn out or say no often enough that the mentoring pipeline degrades.

If you’re short on time

Mentoring is relationship-driven and longer-term. Coaching is situation-driven and shorter. Sponsorship is advocacy in rooms the mentee is not in, and it is the more under-supplied role at the senior level. Hold a portfolio of mentor-mentee relationships rather than a single one, time-box each relationship explicitly, and treat the work as recognised in performance reviews. The two highest-burning failure modes are accepting more requests than you can serve and accepting requests when sponsorship would have been more valuable.

Why this matters at the Council level

The Council’s mandate includes succession. The next cohort of senior engineers does not appear by accident; senior contributors notice the engineers operating at the level, mentor them through harder work, and recommend them for invitation. Without active mentoring at the Council level, the senior-IC pipeline starves and existing senior engineers burn out.

Reilly’s “Being Glue” essay frames the underlying problem directly: mentoring is real work, and if the company does not recognise it in performance reviews, it gets done disproportionately by people whose other contributions are also under-recognised. The Council is the layer where this recognition either lands or does not land. A Council that treats mentoring as recognised work creates a sustainable mentoring culture; a Council that lets it go unrecognised reproduces the failure mode at scale.

Three modes: mentoring, coaching, sponsorship

The literature distinguishes three. Each is a different shape of senior-IC time. Each is needed at different moments. Conflating them produces poor outcomes for both sides of the relationship.

Mentoring. Relationship-driven, longer-term. A mentor and mentee meet on a cadence (typically monthly), discuss the mentee’s evolving career and current work, and the mentor offers perspective grounded in experience. The relationship has shape: it begins, it has a duration, it ends.

Coaching. Situation-driven, shorter, more directive. A coach addresses a specific situation the coachee is in. “I have an RFC to write and I have not done one before” is a coaching situation. A coaching engagement is bounded to the situation; once the situation resolves, the engagement ends.

Sponsorship. Advocacy in rooms the sponsee is not in. A sponsor recommends the sponsee for opportunities, vouches for the sponsee’s work, and uses their political capital to accelerate the sponsee’s career. Drawn from Hewlett’s Forget a Mentor, Find a Sponsor (HBR Press, 2013); Reilly is direct that at the senior level, sponsorship is more valuable and more under-supplied than mentoring.

The three relate. A long-term mentoring relationship may include moments of coaching and acts of sponsorship. A coaching engagement may, if successful, lead to a sponsorship relationship later. But the modes are different enough that conflating them confuses both parties.

A senior contributor evaluating a request from a junior engineer is helped by naming which mode the request implies. “Can I get your perspective on my career direction over the next year?” is a mentoring request. “I need to write my first RFC by Friday” is a coaching request. “I am up for a Staff promotion and I would like you to advocate for me” is a sponsorship request. Different responses are appropriate to each.

The mentor portfolio

Hogan’s central frame in Resilient Management. A senior engineer should not have a single mentor; they should have a small portfolio of mentors covering different facets of their work and life.

Why a portfolio rather than one. A single mentor’s career shape may diverge from the mentee’s. The mentor who was an exceptional infrastructure engineer may be the wrong mentor for a mentee moving into AI-platform work. A mentor who is excellent on technical questions may be the wrong mentor for political ones. A mentor who is at peer level professionally may be the wrong mentor for executive-shaped career conversations.

A portfolio addresses this. The mentee names the facets where they want perspective (technical depth in their domain, political navigation, career-shape decisions, well-being and sustainability) and identifies a mentor for each. Each mentor relationship is lighter than a single all-purpose mentor would be, but the portfolio covers more ground.

For Council members, the portfolio frame matters in two directions. First, Council members are themselves mentees in someone’s portfolio (they have peers and seniors who they consult on career-shape decisions). Second, Council members appear in others’ portfolios, typically as the technical-depth mentor for their domain. Naming the position explicitly helps both sides: the mentor knows which facet they are covering, and the mentee knows where to go for the other facets.

Time-boxing

Mentoring relationships have a stated duration. Hogan is direct that this is the single most important mechanism for preventing the burnout pattern.

A typical mentoring engagement runs a quarter or two. The mentor and mentee agree at the start: “We will meet monthly for two quarters, then evaluate whether to continue.” At the end of the duration, they have an explicit conversation about whether to extend, change shape, or close. Closing is not a failure; it is the normal end of a relationship that has served its purpose.

Without the time-box, two failure modes emerge. The mentor accumulates a portfolio of indefinite relationships and either ghosts mentees or burns out. The mentee assumes the relationship continues forever and is hurt when the mentor’s attention drifts.

The time-box also makes saying yes easier. A senior contributor who can offer a quarter of monthly conversations can plausibly fit one more mentee into their portfolio; a senior contributor who feels they must commit indefinitely cannot.

Coaching engagements have an even tighter time-box. A coaching engagement around a specific situation runs the duration of the situation; usually two to six weeks. Sponsorship is more open-ended but is also lower-touch; sponsorship is showing up in the right rooms with the right framing, not an ongoing time commitment.

Recognising mentoring as performance work

Reilly’s “Being Glue” essay is the canonical text. The argument is simple and consequential: mentoring is real work; if the organisation does not recognise it in performance reviews, the work gets done by people whose other contributions are also under-recognised, and the failure compounds along existing inequities.

The Council’s role here is structural. The Council can recommend, in the charter and in the leadership liaison conversations, that the engineering org’s performance review framework recognises mentoring contributions. The Council cannot mandate this; it sits outside the management chain. But the Council can make the recommendation publicly, contribute to the framework, and surface mentoring contributions in its own minutes.

For Council members specifically, mentoring contributions show up in the annual charter revision when the Council reviews who has been doing what kind of work. A Council member who is mentoring three Black Belt candidates and sponsoring two engineers for promotions has a different shape of contribution from a Council member who is sponsoring two RFCs. Both are legitimate; making both visible is the point.

A framework for accepting versus declining

The literature is genuinely thin on this question. Larson and Reilly both gesture at “be honest about your capacity” without giving an operational answer. Senior contributors who do not have a framework default to either accepting everything (and burning out) or accepting case-by-case based on personal connection (which reproduces existing networks and excludes engineers without those connections).

The framework this chapter constructs has four questions. They are intended as a structure for the senior contributor to think with, not a rubric to share with the asker.

1. What mode is the request implying? Mentoring, coaching, or sponsorship. The mode determines the time commitment and the response shape. A coaching request can usually be accommodated even when the senior contributor’s mentoring portfolio is full.

2. Is this person in your domain or adjacent to it? A senior contributor’s mentoring is most useful when the mentor has direct experience of the work the mentee is doing. A mentor outside the mentee’s domain can offer perspective but cannot offer judgement on technical questions. If the request is in the mentor’s domain, accept is more likely. If the request is far outside, the mentor may serve the mentee better by suggesting a different mentor.

3. Do you have capacity for the time the relationship requires? A mentoring portfolio of three to five active relationships is the working ceiling for most senior contributors. If the portfolio is at the ceiling, a new mentoring relationship requires closing or pausing an existing one, or declining the new request. Accepting beyond the ceiling produces the failure mode where every relationship gets less than the time it deserves.

4. Would sponsorship be more valuable than mentoring here? Sometimes the engineer asking for mentoring is actually asking for sponsorship in disguise. They want their work to be more visible, they want advocacy in promotion conversations, they want introductions to senior people. A sponsoring relationship is lower-touch than mentoring and may serve them better. Naming the question to the asker (without conflating the modes) sometimes redirects the request usefully.

The framework leads to four typical responses:

• Accept as a time-boxed mentoring relationship. “I can offer monthly conversations for two quarters, focused on [their named question]. Let’s check in at the end on whether to continue.”
• Accept as a coaching engagement. “I can pair with you on [the specific situation] for the next four weeks. After that, the relationship closes unless we both name a continuation.”
• Accept as sponsorship. “I am not the right mentor here, but I will advocate for you in [the named rooms]. Send me a paragraph on what you are working on so I can carry the context.”
• Decline with a referral. “I am at capacity for new mentoring relationships this quarter, but [named other senior contributor] has the right context and may have space. I am happy to make the introduction.”

The framework is not a script. The senior contributor uses it to think; the response is in their own voice.

Common failure modes

Drawn directly from the literature.

The unrecognised glue worker. Reilly’s “Being Glue” essay names this most clearly. Mentoring done invisibly is mentoring that destroys the mentor’s career. The mentor’s other contributions are under-counted because so much of their time goes to invisible work. Fix: the recognition pattern above. The Council surfaces mentoring contributions in performance review conversations and the charter recommends that the framework recognises this work.

The collected mentees. Hogan’s blog. A senior contributor who says yes to every request accumulates a portfolio they cannot serve, then either ghosts mentees or burns out. Fix: the four-question framework; the mentoring portfolio ceiling.

Mentor as shadow manager. Fournier’s The Manager’s Path is direct on this. A mentor who begins making decisions that belong to the mentee’s manager produces triangulation: the mentee plays the manager and the mentor against each other. Fix: the mentor stays out of the mentee’s reporting line; performance feedback flows through the manager; the mentor offers perspective, not directive.

The single-mentor monoculture. Hogan. A mentee who models themselves entirely on one senior contributor reproduces that contributor’s blind spots. The mentee’s career narrows. Fix: the portfolio frame; the mentor encourages the mentee to find other mentors for facets the current relationship does not cover.

Mentoring confused with sponsorship. Reilly. Advice without advocacy is unhelpful at the senior level. A mentee who needs sponsorship gets mentoring that fails to move their career. Fix: the four-question framework’s question 4; explicit conversations about which mode the relationship is in.

The all-volunteer Council mentoring corps. A pattern that emerges when the Council is treated as the mentoring body for the whole engineering org. Council members exhaust themselves serving every junior engineer. Fix: the Council mentors at the Black Belt and Council-candidate level. Mentoring at junior levels happens through team managers, individual engineers, and structured mentoring programs that the Council can recommend but does not staff itself.

Sponsorship that becomes patronage. A sponsor who repeatedly advocates for the same small group of engineers, while not advocating for others with comparable work, reproduces a closed network. Fix: the sponsor consciously expands the set of engineers they sponsor; the Council’s annual review surfaces patronage patterns.

Time-box drift. Mentoring relationships that were meant to be a quarter become indefinite. The mentor’s portfolio quietly fills. Fix: the time-box is named at the start and honoured at the end; renewing is a deliberate conversation, not a default.

What this is not

Not a substitute for line management. The mentor is not the manager. Performance feedback, career architecture decisions, and resourcing decisions belong to the manager. The mentor offers perspective and coaching; the manager owns outcomes for the engineer.

Not a closed circle. Mentoring at the Council level is not reserved for engineers who are already part of the senior-IC community. Council members mentor Black Belt candidates, engineers from outside their immediate domain, and engineers whose work is currently under-recognised. The framework helps avoid accidental closure.

Not free of conflict of interest. A mentor who is also a peer reviewer, an RFC approver, or a hiring decision-maker in the mentee’s path has overlapping interests. Naming the overlap explicitly is part of the relationship; pretending it is not there produces worse outcomes for both sides.

Not unlimited. A senior contributor’s mentoring capacity is finite. The portfolio ceiling is not an arbitrary number; it is the real number of relationships any one person can serve well. Pretending otherwise produces the burnout pattern at scale.

Not the same as office hours. Office hours (B.12 in Black Belt) are a parallel surface that scales differently. Office hours are open, time-bounded, and address specific blockers in a session. Mentoring is closed, longer-term, and addresses career or evolving situations across multiple sessions. Both are valuable; conflating them confuses both parties.

Self-check for senior contributors

Adapted from the literature’s common patterns.

Healthy. I have a mentoring portfolio of three to five relationships. Each is time-boxed. I sponsor more engineers than I mentor. Mentoring contributions show up in my own performance review conversations. I decline mentoring requests when my portfolio is full and refer to other mentors.

At-risk. I have a mentoring portfolio of seven to ten relationships. Several are indefinite. I cannot reliably remember the last conversation I had with at least two of my mentees. Mentoring is not visible in my own performance conversations.

Unhealthy. I have lost count of my mentoring relationships. I ghost mentees or cancel sessions repeatedly. I cannot tell whether I am sponsoring anyone. My burnout signals are showing.

The Council surfaces unhealthy patterns at the annual charter revision. The fix is structural, not exhortatory: the time-box discipline, the portfolio ceiling, the recognition pattern.

What you can say after this chapter

“I distinguish mentoring from coaching from sponsorship and choose appropriately for each request. I hold a time-boxed portfolio of mentoring relationships within a working ceiling. I sponsor engineers actively and treat sponsorship as the more under-supplied role at the senior level. I name mentoring contributions in performance conversations because the literature is unambiguous that unrecognised mentoring work compounds existing inequities.”

Where to go next

C.5 covers external voice: writing, speaking, OSS contributions, and the personal-versus-company brand tension. Council members represent the program externally; the chapter walks the patterns and the governance.

Previous: ← C.3 The RFC pipeline · Next: → C.5 External voice

Further reading

• Lara Hogan, Resilient Management (A Book Apart, 2019). The mentor portfolio frame.
• Tanya Reilly, “Being Glue” (2019). The most-cited piece on under-recognised mentoring work.
• Tanya Reilly, The Staff Engineer’s Path (O’Reilly, 2022), the chapter on “Levelling Up”.
• Will Larson, lethain.com, posts on coaching versus mentoring.
• Camille Fournier, The Manager’s Path (O’Reilly, 2017). The management-side counterpoint.
• Sylvia Ann Hewlett, Forget a Mentor, Find a Sponsor (HBR Press, 2013). Upstream source on sponsorship.
• Lara Hogan’s blog on burnout, time-boxing, and the feedback equation.