Cabo Advisory

← Insights··Updated

From founder-led to professionally managed: the governance transition

Founders are replaced early and often, and yet founder involvement is associated with outperformance and innovation. The evidence is not contradictory; it is about two different things. The transition is not "founder out, manager in" but a re-allocation of three assets — authority, information and legitimacy — so that the founder's innovation function stays inside a system of accountability. And the same design lands on different legal footings in Stockholm, London and Madrid.

The wrong question

On 30 September 2025 Spotify announced that Daniel Ek would become executive chairman from 1 January 2026, with Gustav Söderström and Alex Norström — co-presidents since 2023 — becoming co-chief executives, reporting to him. Spotify described the role as "a European Chairman setup" in which Ek "will determine capital allocation"; Ek himself said he would "focus on the long arc of the company".1 It is the most visible recent example of a transition that most successful technology companies eventually face, and it illustrates the point that the question boards usually ask — should the founder stay chief executive? — is the wrong one.

The statistics that frame the debate are two decades old and still cited daily. In Noam Wasserman's study of 212 American start-ups, "by the time the ventures were three years old, 50% of founders were no longer the CEO; in year four, only 40% were still in the corner office", and fewer than 25% led their company's initial public offering.2 Against that, Bain's research on the "founder's mentality" found that companies where the founder is still involved delivered "3.1 times higher total shareholder returns over the past 15 years, compared to others" — and the wording matters: involved, not necessarily chief executive.3 The academic literature points the same way. Fahlenbrach found that founder-led firms among the largest US public companies earned benchmark-adjusted returns of 8.3% a year, 4.4% after controls, and invested more in research and capital expenditure.4 Lee, Kim and Bae, using the sudden deaths of chief executives as a natural experiment, found that an exogenous switch from a founder to a professional CEO "is associated with a 43.8% decrease in a firm's citation-weighted patent count, even when controlling for R&D expenditures".5

Read together: the founder's value is real and concentrated in innovation and capital allocation; the founder's tenure as CEO is usually short; and three-quarters of departing founder-CEOs stay involved as chair, advisor or investor.6 The right question is what the company's decision system looks like after the founder stops being the decision system.

What actually transfers: authority, information, legitimacy

A founder-led company runs on three things the founder holds personally. Formal authority — the signing powers, the hiring decisions, the budget. Information — the founder knows the product, the customers and the engineers without needing a report. And legitimacy — people do what the founder says because it is the founder saying it.

An org-chart change moves the first and leaves the other two where they were. That is the source of the "shadow CEO" problem, in which a founder-chair retains the informal reporting lines while a professional CEO holds the title. It is also what Paul Graham's "Founder Mode" essay of September 2024 was describing from the other side. Of the conventional advice to "hire good people and give them room to do their jobs", Graham wrote that "what this often turns out to mean is: hire professional fakers and let them drive the company into the ground"; his answer was to legitimise skip-level engagement by founders.7 Founder mode is an operating model, not a governance model; it says nothing about who holds the CEO to account. A transition that works has to move all three assets deliberately, and the governance instruments exist to do it.

Stage 1: codify before you change people

The first stage happens before anyone's title changes. Write down how the board works and what the chief executive may decide alone. Both the Swedish and UK codes require it: the Swedish Corporate Governance Code provides that "the formal procedures for the work of the board of directors are to be stipulated by the board" and are "to be clear and well documented",8 and the UK Corporate Governance Code requires that "the responsibilities of the chair, chief executive, senior independent director, board and committees should be clear, set out in writing, agreed by the board and made publicly available".9

The statutory hooks for a delegation-of-authority matrix are already there. Under the Swedish Companies Act, the board "is responsible for the company's organisation and the management of its affairs", may appoint a managing director, and the managing director "shall manage the day-to-day management in accordance with the board's guidelines and instructions" — which makes the board's written instruction to the CEO the legal boundary of the CEO's job.10 Under the Spanish Ley de Sociedades de Capital, the board may delegate powers to one or more consejeros delegados "establishing the content, limits and modalities of the delegation", and article 249 bis lists what can never be delegated: general policies and strategy, authorising conflicts of interest, formulating the annual accounts, appointing and removing the consejeros delegados and the executives who report directly to the board or any of its members, director remuneration and calling the general meeting.11

The practical work is alignment. A scale-up typically runs three inconsistent lists of "big decisions": the reserved matters in the shareholders' agreement, the board's own rules of procedure, and the statutory non-delegable powers. The 2023 revision of the BVCA model documents "tidied up and made clearer" the reserved-matters schedule and distinguished investor-director consent for operational matters from investor consent for reserved matters;12 the same discipline — one list, one hierarchy, one place where each decision lives — is the foundation on which a professional CEO can be given real authority.

Stage 2: rebuild the board around independence, not representation

Founder-led boards are usually assembled from the people who funded the company. That is representation, not governance, and it is the structure that failed at Northvolt: a founder-CEO surrounded by strategic and financial shareholder representatives, Volkswagen alone holding a 20% stake and a board seat after a €900 million investment, and no evident independent challenge on capital allocation and expansion pace. The company filed for Chapter 11 on 21 November 2024 "stating it had only $30 million in cash, enough for them to operate one more week, while carrying $5.8 billion in debts"; its founder-CEO resigned the next day, and in March 2025 the company filed for bankruptcy in Sweden.13 Klarna, in October 2024, showed the other failure mode of a representative board, with seven of eight directors voting to remove an ally of a co-founder while the company was preparing an IPO — a board fracturing along shareholder-bloc lines rather than deciding as a board.14

The codes describe the alternative precisely. Under the Swedish Code, "no more than one elected member of the board may be a member of the executive management", a majority of the shareholder-elected directors "are to be independent of the company and its executive management", and at least two of those are also to be "independent in relation to the company's major shareholders".8 Under the UK Code, the chair "should be independent on appointment", at least half the board excluding the chair should be independent non-executives, and the chair "should not remain in post beyond nine years".9 Neither code binds a private scale-up. Both describe what its next buyer, lender or listing advisor will expect to see, and investors routinely write the substance into shareholders' agreements.

Stage 3: move the founder deliberately — the executive-chair construction

Spotify's design — founder as executive chairman with strategy and capital allocation, operators as CEO — is the current reference model, and it is legally textured in ways that differ by jurisdiction.

In Sweden, a public company's chair "may not be the managing director of the company" — a statutory prohibition, not a code recommendation.10 Where the chair nonetheless has executive duties, the Code requires that "the division of work and responsibilities between the chair and the chief executive officer is to be clearly stated in the board's statutory rules of procedure and its instruction to the chief executive officer".8 In the United Kingdom the Code is explicit that "the roles of chair and chief executive should not be exercised by the same individual" and that "a chief executive should not become chair of the same company"; a founder-CEO who moves to the chair is a departure that has to be explained, and it remains common — Spencer Stuart found that 38 percent of S&P 500 executive chairs were the outgoing CEO.915 In Spain, a director given executive functions must have a contract with the company approved by two-thirds of the board, from whose deliberation the director "must abstain", setting out "all the items for which he or she may obtain remuneration"; the director "may not receive any remuneration" outside it.11 The Spanish rule is, in effect, a mandatory transparency mechanism for the founder-chair's package.

Two further instruments are recalibrated at the moment of the role change. The first is the founder's equity: vesting terms written for a founder-CEO rarely fit a founder-chair, and leaver provisions written to protect investors from a departing executive can punish a founder who steps back voluntarily. The February 2025 edition of the BVCA model documents removed resignation as a bad-leaver trigger precisely because resignation "is rarely a Bad Leaver event".16 The second is the founder's information rights: a founder who is no longer an executive should receive information as a director, through the board, not through the reporting lines the CEO now owns. A related problem — founders who remain after a sale with an earn-out — is treated in Part 4 of this series.

Stage 4: re-wire information and controls

The transition fails quietly when the professional CEO has the authority but the board still lacks the information. The instruments here are unglamorous: tiered information rights (the BVCA model documents have offered a major-investor-only information-rights option since 2023, and since February 2025 an optional power to restrict information flow to strategic investors),1216 a management reporting cadence agreed by the board rather than inherited from the founder, and a private-company version of the UK Code's new Provision 29, under which boards declare the effectiveness of their material controls and describe "any material controls which have not operated effectively".9

The Swedish Code adds one device that every founder transition should adopt whatever the jurisdiction: "the board is to continuously evaluate the work of the chief executive officer" and examine it "formally at least once a year, and no member of the executive management is to be present during this evaluation process".8 An annual evaluation of the CEO with the founder-chair in the room and the CEO outside it is the formal counterweight to the founder's informal legitimacy.

There is a liability dimension too. Spanish law protects directors' business decisions only where the director "has acted in good faith, without a personal interest in the matter, with sufficient information and in accordance with an appropriate decision-making procedure".17 The documentation of the transition — the instruction to the CEO, the information the board received, the minutes of its evaluation — is also the directors' defence.

Three failure modes

The shadow CEO. The founder-chair keeps the informal reporting lines; engineers and key customers go to the founder; the professional CEO learns of decisions after they are made. The cure is Stage 1 and Stage 4: a written division of labour, and information that flows through the board.

The orphaned CEO. The professional CEO's authority is hollowed out by investor consents on one side and founder vetoes on the other. Uber's 2017 crisis was the mirror image — a founder who had negotiated control of additional board seats in a 2016 charter amendment, which investors then sued to unwind after he had resigned as CEO under pressure.18 Control instruments granted in a funding round become the battleground when the transition is contested; audit them before the transition, not during it.

The board as referee. OpenAI's board removed its chief executive in November 2023 on the ground that he "was not consistently candid in his communications with the board", and reinstated him within days. The independent review concluded that there had been a "breakdown in trust", that the board had acted "on an abridged timeframe, without advance notice to key stakeholders", and that "his conduct did not mandate removal"; the remedies were process instruments — new governance guidelines, a strengthened conflict-of-interest policy, a whistleblower hotline and new committees.1920 A board with the legal power to remove a founder but without process legitimacy loses to the founder's coalition.

Plan for the second transition

The professional CEO is not the end state. Russell Reynolds recorded 234 chief-executive exits across thirteen major indices in 2025, an eight-year high, and found that 86% of appointees were first-time CEOs;21 Spencer Stuart found that 42% of departing European CEOs in 2024 had served under five years;15 and in private-equity-backed companies 65% of firms report changing the CEO during the hold.22 The founder-to-professional transition should therefore be written to survive the professional's own departure: the CEO's succession plan, the founder's return path (Starbucks, where Howard Schultz twice returned as chief executive, is the familiar example), and the board's ability to run a process it can defend all belong in the same instruments.

A one-page instrument checklist

  • Board rules of procedure and a written CEO instruction — Swedish Code, chapter 7; UK Code Provision 14; ABL 8:29; LSC art. 249.1.
  • One reconciled list of reserved and non-delegable matters — shareholders' agreement, board rules, LSC art. 249 bis.
  • Board composition targets — Swedish Code 4.3–4.5; UK Code Provisions 9, 11 and 19.
  • Executive-chair terms — Swedish Code 6.2 division of labour; UK Code Provision 9 explanation; LSC art. 249.3–4 contract, two-thirds approval and abstention.
  • Founder equity and leaver terms re-cut for the new role — BVCA February 2025 leaver changes as the reference point.
  • Information rights and a controls declaration — tiered rights; a Provision 29-style statement.
  • Annual CEO evaluation without executives present — Swedish Code 9.2.
  • Succession and return paths written down — for the professional CEO and for the founder.

The founder who moves to the chair with strategy and capital allocation, an operator who runs the company with real authority, and a board that can evaluate both is a design that the evidence supports and that three legal systems accommodate. What none of them supplies is the willingness to write it down before the crisis. That part is the board's job — and, as Part 1 of this series argues, it is what the company's owners actually need from it.

Related reading in this series

Disclaimer

This article is general commentary for professional readers. It is not legal, tax or financial advice, does not address the circumstances of any particular company, and does not create an advisor–client relationship with Cabo Advisory SL. Statutory provisions and governance codes are summarised, not reproduced, and are stated as at the date of writing; take advice on the law of the relevant jurisdiction before acting.

Notes and sources

Footnotes

  1. Spotify Newsroom, "Spotify Announces Leadership Evolution: Daniel Ek to Become Executive Chairman, Alex Norström and Gustav Söderström to Become Co-CEOs in January 2026", 30 September 2025. https://newsroom.spotify.com/2025-09-30/spotify-announcement-daniel-ek-executive-chairman/ ↩

  2. Noam Wasserman, "The Founder's Dilemma", Harvard Business Review, February 2008 (reprint R0802G); text as reprinted in HBR's 10 Must Reads on Entrepreneurship and Startups. https://hbr.org/2008/02/the-founders-dilemma ; https://www.oreilly.com/library/view/hbrs-10-must/9781633694392/Text/chapter011.html ↩

  3. Bain & Company, "85 percent of company shortfalls in achieving sustained, profitable growth are caused by internal breakdowns, not external factors" (press release), 21 July 2016, https://www.bain.com/about/media-center/press-releases/2016/internal-breakdowns-cause-85-percent-of-company-shortfalls-in-achieving-profitable-growth ; Chris Zook and James Allen, "Barriers and Pathways to Sustainable Growth: Harnessing the Power of the Founder's Mentality", Bain & Company, 19 July 2016, https://www.bain.com/insights/founders-mentality-barriers-and-pathways-to-sustainable-growth/ ↩

  4. Rüdiger Fahlenbrach, "Founder-CEOs, Investment Decisions, and Stock Market Performance", Journal of Financial and Quantitative Analysis 44(2), April 2009, pp. 439–466. https://ideas.repec.org/a/cup/jfinqa/v44y2009i02p439-466_09.html ↩

  5. Joon Mahn Lee, Jongsoo Kim and Joonhyung Bae, "Founder CEOs and innovation: Evidence from CEO sudden deaths in public firms", Research Policy 49(1), February 2020. https://www.sciencedirect.com/science/article/abs/pii/S0048733319301817 ↩

  6. Egon Zehnder, "Founder Vs CEO Dilemma – Knowing When to Pass the Torch", April 2025 (tracking 52 founder-CEO departures in the first half of 2023). https://www.egonzehnder.com/functions/chief-executive-officers/founders/insight/headline-riding-the-founder-succession-wave-why-top-ceos-are-passing-the-torch-in-2023 ↩

  7. Paul Graham, "Founder Mode", September 2024. https://paulgraham.com/foundermode.html ↩

  8. Swedish Corporate Governance Board, The Swedish Corporate Governance Code, applicable from 1 January 2024, rules 4.3, 4.4, 4.5, 6.2 and 9.2 and the introduction to chapter 7. https://bolagsstyrning.se/Userfiles/Koden/Dokument/Eng/swedish_corp_governance_code_applicable_from_1_january_2024_redline.pdf ↩ ↩2 ↩3 ↩4

  9. Financial Reporting Council, UK Corporate Governance Code 2024 (published 22 January 2024; applying to financial years beginning on or after 1 January 2025, Provision 29 from 1 January 2026), Principle G and Provisions 9, 11, 14, 19 and 29. https://www.frc.org.uk/library/standards-codes-policy/corporate-governance/uk-corporate-governance-code/ ↩ ↩2 ↩3 ↩4

  10. Aktiebolagslag (2005:551), 8 kap. 4, 27, 29 and 49 §§ (consolidated text, Sveriges riksdag); quotations in the text are our translations of the Swedish. https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/aktiebolagslag-2005551_sfs-2005-551/ ↩ ↩2

  11. Real Decreto Legislativo 1/2010, de 2 de julio, Ley de Sociedades de Capital, arts. 249 and 249 bis (as amended by Ley 31/2014); quotations in the text are our translations of the Spanish. https://www.boe.es/buscar/act.php?id=BOE-A-2010-10544 ; article texts as reproduced at https://www.iberley.es/legislacion/articulo-249-ley-sociedades-capital and https://www.iberley.es/legislacion/articulo-249-bis-ley-sociedades-capital ↩ ↩2

  12. Alex Potop and Tom Watkins, "Faster and more cost-effective: unpacking the BVCA's model document updates", DLA Piper, 17 July 2023. https://www.dlapiper.com/en-nz/insights/publications/2023/07/faster-and-more-cost-effective-unpacking-the-bvcas-model-document-updates ↩ ↩2

  13. "Northvolt", Wikipedia (accessed 26 September 2026), https://en.wikipedia.org/wiki/Northvolt ; Matthew Lynas, "Northvolt CEO steps down following bankruptcy filing", pv magazine, 22 November 2024, https://www.pv-magazine.com/2024/11/22/northvolt-ceo-steps-down-following-bankruptcy-filing/ ↩

  14. Connie Loizos, "Klarna is about to oust another board member", TechCrunch, 6 October 2024. https://techcrunch.com/2024/10/06/klarna-is-about-to-oust-another-board-member/ ↩

  15. Spencer Stuart, "2024 CEO Transitions: The measure of the market", February 2025, https://www.spencerstuart.com/research-and-insight/2024-ceo-transitions ; Spencer Stuart, 2024 CEO Transitions in Europe, February 2025, https://www.spencerstuart.com/-/media/2025/03/2024-ceo-transitions-europe/2024-ceo-transitions-europe.pdf ↩ ↩2

  16. UK Private Capital (formerly BVCA), "BVCA Model Documents for Early Stage Investments: key changes to February 2025 edition", 26 February 2025. https://www.ukprivatecapital.co.uk/resource/bvca-model-documents-for-early-stage-investments-key-changes-to-february-2025-edition.html ↩ ↩2

  17. Ley de Sociedades de Capital, art. 226.1 (protección de la discrecionalidad empresarial); quotation is our translation of the Spanish. https://www.iberley.es/legislacion/articulo-226-ley-sociedades-capital ↩

  18. Kate Vinton, "Benchmark Capital Sues Former Uber CEO Travis Kalanick For Fraud", Forbes, 10 August 2017. https://www.forbes.com/sites/katevinton/2017/08/10/benchmark-capital-sues-former-uber-ceo-travis-kalanick-for-fraud/ ↩

  19. OpenAI, "OpenAI announces leadership transition", 17 November 2023. https://openai.com/index/openai-announces-leadership-transition/ ↩

  20. OpenAI, "Review completed & Altman, Brockman to continue to lead OpenAI", 8 March 2024. https://openai.com/index/review-completed-altman-brockman-to-continue-to-lead-openai/ ↩

  21. Russell Reynolds Associates, "Global CEO Turnover Index" (full-year 2025 data; accessed 26 September 2026). https://www.russellreynolds.com/en/insights/reports-surveys/global-ceo-turnover-index ↩

  22. AlixPartners, "Nearly two-thirds of private equity firms replace portfolio company CEOs during the holding period, AlixPartners survey finds" (press release for the 11th Annual Private Equity Leadership Survey), 25 March 2026. https://www.alixpartners.com/newsroom/alixpartners-11th-annual-private-equity-leadership-survey/ ↩