IF Jes Staley becomes new chief executive of Barclays, he is bound to ask what the bank’s staff think. Aides will probably hand him the ‘colleague engagement’ survey, one of many gauges his predecessor Antony Jenkins set up to check whether the bank’s culture had improved.

But will it help? In-house staff questionnaires are balm to worried bosses, who are insulated from the truth by two facts: first, in any company above a certain size, they are far from shop floor gossip; second, eager acolytes mostly tell them what they want to hear.

Internal polls can also deflect nosy reporters’ questions. Six months before it abandoned its independent turnaround strategy and sold its devices business to Microsoft, Nokia told me internal measures of staff satisfaction were rising — much good it did the staff later laid off. At AB InBev, insiders confided concerns about the brewer’s cult-like management approach, but officials pointed me to improving levels of worker engagement.

Outside agencies’ assessments of workplace contentment are even more valuable public relations fodder. Barely anyone beyond local media covered last week’s 2015 ranking of the world’s best multinational workplaces, but half a dozen of the companies on the list published press releases drawing attention to their audited niceness.

The overall message is that science trumps anecdote. But even fans of staff polls say they can be skewed by self-selection (because disgruntled staff leave) or self-delusion (because the remaining staff have to convince themselves they were right to stay).

The sheer size of some modern companies creates further noise. Few multinational workplaces offer a uniform staff experience that can be summed up in a single survey

The sheer size of some modern companies creates further noise. Few multinational workplaces offer a uniform staff experience that can be summed up in a single survey. Google tops the multinationals list, and has done for three years. But it is hard to tell whether it scores highly because it has the pick of the best and most highly motivated staff in technology, or whether it has the pick of those staff because it scores highly.

Barclays, on the other hand, has contrasting cultures in retail and investment banking, which also vary for different local markets. Mr Staley would inherit a staff engagement measure that slipped from 74pc to 72pc between 2013 and 2014, compared with a 2018 target of at least 87pc. Even with the help of more specific and frequent surveys, he will need help unpicking the data to work out which parts of the operation are holding Barclays back.

Staff surveys increasingly have to overcome another hurdle: the very definition of ‘staff’.

In other bank news last week, HSBC forced a pay cut and an unpaid holiday on contractors in its investment banking division. For large networked companies, the contractors, part-timers and workers drawn from the human cloud are the staff.

Rob Goffee and Gareth Jones’s new book Why Should Anyone Work Here? is peppered with survey-style questions to help readers work out how to create an attractive workplace with engaged workers. But, as they point out, ‘building a cohesive, meaningful team — let alone corporate community’ from a fragmented group of loosely connected short-term labourers is demanding. Polling them accurately will be harder still.

When I asked an experienced director of three large multinationals whether her companies paid much attention to staff surveys, she looked astonished that I doubted it. Directors at each of her companies pore over the data that you and I bash into online forms in 10 rushed minutes over a sandwich at our desks.

At one group where she was a non-executive, an employee poll revealed a critical flaw that led to the sacking of a manager before he could do more damage. At another, a jump in the staff sentiment gauge reassured the board it had taken the right strategic decision. Provided you look at the trends, and compare them with companies that are your peers, such surveys can be invaluable, this director said.

I remain a sceptic (but then FT journalists are regularly chastised for being slow to fill in our periodic engagement survey). Yet even I am coming round to the idea that companies are right to cling to these blunt tools. Even with all their drawbacks, questionnaires are the least bad way for directors to find out what people they are never likely to meet actually think. Now they just need to act on what they learn.

andrew.hill@ft.com

Twitter: @andrewtghill

Published in Dawn, Business & Finance weekly, October 26th , 2015

On a mobile phone? Get the Dawn Mobile App: Apple Store | Google Play