Co-CEOs don't always see eye to eye about which way a business should go. Photo / Getty Images
Last week's award of the Booker Prize to two winners upset almost everybody except the laureates themselves, Margaret Atwood and Bernardine Evaristo.
The decision — or rather the indecision — distressed me, too, and not only because it set a bad example for the Financial Times and McKinsey book award,which I help to organise (judges, if you're reading: don't even think about it). When a prize or a leadership position is split, it usually means someone has shirked their duty to make their mind up.
At least Evaristo and Atwood were not asked to run something together, as co-chief executives are. Whether decisions to share leadership are taken for good, or bad, reasons, the outcome is usually the same: one executive clings on and the other slinks away.
Co-CEO arrangements, perhaps as a result of such experiences, seem to be rarer than they used to be, but there has been a recent flurry of joint appointments.
WeWork last month named Artie Minson, the group's long-serving chief financial officer, and Sebastian Gunningham, its vice-chairman, to replace founder Adam Neumann after the humiliating withdrawal of the co-working group's planned share listing. SAP, the German software company, installed Jennifer Morgan and Christian Klein as co-chief executives in place of Bill McDermott last week.
SAP is the exception that proves the rule that boards thinking of naming joint leaders should think again. Its founders shared the chief executive role and so did Mr McDermott before he took sole control in 2014.
Most other examples are unsatisfactory for one reason or another. For instance, Deutsche Bank's current woes can be traced back to the handover in 2012 from Josef Ackermann to Jürgen Fitschen and Anshu Jain, who lasted three years before low returns, lacklustre strategy and legal wrangles undid them.
Even Deutsche's awkward compromise was better than the jerry-rigged co-CEO structures that companies often throw up after "mergers of equals". When I wrote about these doomed love stories earlier this year, a knowledgeable reader pointed out such solutions were almost always about cranking up the pay and bonuses of the existing leaders: "It's a way of buying the support of a CEO that doesn't want to leave . . . It has everything to do with individual rewards over the short-term and nothing to do with operating efficiency or shareholder rewards".
The prospects for pairings set up with the best intentions are not much better. One frequently heard justification to divvy up the leadership is that in a complex world, it makes sense to distribute the burden. This is where the saying "a problem shared is a problem halved" comes up against that other proverb about too many cooks spoiling the broth.
Declan Fitzsimons of Insead, who has studied shared leadership in organisations, warns that if you multiply the number of leaders, you will also multiply the number of relationships their underlings have to manage, as they answer not just to each individual but to the two co-leaders as a pair. "Team members are discombobulated: 'Who do we relate to? Who's in charge here?'"
Attempts to split responsibilities, as WeWork has — corporate functions such as finance to Mr Minson; product, design and sales to Mr Gunningham — fall foul of the fact that collaboration and information usually cut across corporate departments.
Adding to the potential for disaster, a 2017 study found that dysfunctional paranoia and power struggles were worse, the higher the rank assigned to co-leaders. On that basis, I dread to think what must be happening at Azimut, the Italian asset manager, which earlier this year named five chief executives of the group, arguing that each had "a proven record of major accomplishments".
When shared leadership fails, Prof Fitzsimons says, powerful group dynamics take over and "it's the unconscious that will run the show". Such an atavistic outcome can be avoided, but it involves hard work.
Companies can learn how to foster productive co-CEO relationships, as SAP seems to have done. The board can prepare pairs of chief executives for their co-operative role, rather than simply launching them into it. Joint chiefs should also try to air any differences in front of their teams, rather than allowing such rifts to fester in private before erupting in public.
All this is easier said than done. So why complicate what is already an unpredictable succession process? Half a prestigious prize is better than none, seemed to be the gracious conclusion of Evaristo and Atwood after the botched Booker decision. In real life, though, such a compromise rarely yields such a happy ending.