After Ferguson

Manchester United lost its manager, chief executive and coaching staff in one summer, and Moyes lasted under ten months. What a manager leaves behind shows whether they were running a system or were the system, and engineering managers almost never get to see it.

In the summer of 2013 Manchester United changed a good deal more than its manager. Alex Ferguson's retirement was announced on 8 May after 26 years; David Gill, the chief executive who had run the club around him for a decade, had said in February that he would go at the end of June; and when David Moyes arrived he released Ferguson's assistant Mike Phelan and goalkeeping coach Eric Steele, while first-team coach René Meulensteen left within weeks. The people who knew how the place worked went out of the door at roughly the same time, and Moyes was sacked on 22 April 2014, 51 games and a little under ten months after he started.

The post-mortems from both sides are more interesting than the result. Ferguson's, in Leading, insists that "the process was perfect" and then, a few lines on, says Moyes should have kept Phelan, "who would have been the invaluable guide to the many layers of the club", because changing routines players are comfortable with "saps morale and immediately provokes players to question the new man's motives." Meulensteen described the setup Moyes inherited as "a well-oiled machine" that Ferguson had expected to keep running. Both men place a large part of what made United work in the staff and the routines rather than in Ferguson himself, and yet the machine didn't survive a season without him, which means either that it was Ferguson all along or that a system needs the people who know where its levers are. This series has been asking what a manager is for, and it ends on this version of the question.

The first few months after a handover are the obvious place to look for an answer, and the evidence says they show almost nothing. Clubs sack managers during slumps, and slumps tend to end on their own. Van Ours and van Tuijl, looking at fourteen Dutch seasons, found that teams performed better after replacing a coach, and so did a control group of comparable teams that kept theirs; Heuer and colleagues found that an in-season dismissal in the Bundesliga has "basically no effect"; in Serie A, De Paola and Scoppa found the apparent gain disappears once you account for the dip before the sacking. The one dissent is a study of Ligue 1 from 1998 to 2018, which found a positive short-run effect that showed up only in home games, which its authors put down to fan pressure rather than a better manager. Van Ours's more recent work adds that clubs in the top five leagues sack on results even when the expected-goals numbers were fine, so the bounce is real but modest, and it mostly measures the scoreboard rather than either manager.

Liverpool ran the same test from the other direction. Arne Slot replaced Jürgen Klopp in 2024 with almost no new players — Federico Chiesa, plus a goalkeeper who stayed on loan at Valencia — and won the league in his first season. A year later, after a record £446.5m summer, Liverpool finished fifth and Slot was sacked on 30 May 2026. Whether that title was the last return on Klopp's system or Slot's own work is something the table can't settle, because the squad was remade before anyone could find out.

Coaching trees are the slower test, since they show what a manager's people can do once they're somewhere else. Guardiola's is winning: Mikel Arteta, his assistant at City, took Arsenal to the 2025/26 title; Vincent Kompany, his former captain, has won two Bundesliga titles at Bayern; and Enzo Maresca, another former assistant, was appointed to succeed Guardiola himself at City in June. The exception is Domènec Torrent, a decade at Guardiola's side, whom Flamengo sacked after ten weeks in 2020. Ferguson's tree is mostly sackings — Meulensteen lasted 16 days at Anzhi and about eleven weeks at Fulham, Phelan was sacked by Hull in 20th place — though not entirely, since Laurent Blanc, who played two seasons for him, won Ligue 1 with Bordeaux and three more with PSG. Survivorship cuts both ways here: Guardiola's alumni were handed Arsenal, Bayern and City, and a tree says as much about who gets hired where as about what anyone learned.

The other half of any succession is the squad, and here I can speak from my own time as a manager rather than from the record. The worst case is being given a team of poor performers that the company can't move on; they usually know the company doesn't like them, and they're happy to hang around. It's rare, but it happens in football too. In August 2024 Maresca, newly at Chelsea, said "I am not working with 42 players. More than 15 players are training apart and not with the team", and was careful to add, of Raheem Sterling, "I am not saying Raheem is not a good player but I prefer different kinds of wingers." A year later Rúben Amorim's United had its own exiled "bomb squad": Marcus Rashford, Jadon Sancho, Alejandro Garnacho, Antony and Tyrell Malacia, who had told the club they wanted to leave. By the end of the window four had gone, and Malacia was left behind when a loan to Elche collapsed. Maresca's line about Sterling is the fair way to read all of them: these were players a new manager had no use for, at clubs that had signed them for someone else's plan.

That's also where the analogy breaks. A football bomb squad is a mismatch with a price attached: there's a market, a loan system and a contract end date, and United cleared four of its five in a single window (the finances of that are a different series). An engineering team has none of those. Nobody buys an engineer's remaining contract, there's no loan to a team where they'd fit better, and "can't move on" usually means the employer won't or can't run the process, so the people stay and both sides know where they stand. The second break is that football runs its succession test in public every summer and engineering barely runs it at all: managers stay for years, their departures go unrecorded, and nobody keeps a table of their successor's first season.

The question Ferguson's last summer raises, whether the team was good because of a system or because of a person, is one most engineering managers will never see answered from the outside, because they won't be watching when their successor takes over. Ferguson's own account puts the answer in the staff and the routines, which is a convenient conclusion for the person who left and an untestable one, since Moyes changed both at once. If you're running a team now, and you can't wait to find out after you've gone, what would you look at while you're still there to tell which one you've built?

No comments yet