CASINO BANKING

Sitting at a table on an early morning train to Manchester, I was intending to use the journey to do some work. That was before three smartly dressed, articulate young compatriots joined me and proceeded to order half-a-dozen cans of cider – which turned out to be their ‘work’ for the trip.

Wondering why the cider at 7am, I clocked that one of the three was wearing playing-card cufflinks. From that I assumed they’d spent the night in a casino and had a hard night battling the odds. I was right about the odds but wrong about the casino.

It turned out that my festive companions worked in the complaints-handling department of a major retail bank, based in Scotland. They were celebrating the end of their last day’s (or rather night’s) work for the week and were on their way home. High pay for anti-social hours enabled all three, who lived in England, to make the weekly commute to Scotland and still take home generous pay. Their motives for the lifestyle varied: one was saving a deposit for a house, another paying off student debt, the third living ‘the high life’.

Their job success hinged on completing three complaints cases a night. Over the course of the first three cans of cider they proceeded to compare notes on their standard methods for meeting the targets with the least possible effort.

How to meet targets 1

The favoured method involved getting to work before their supervisor to allocate themselves their first case for the night. This meant that when the boss arrived, they were already halfway through their (self-chosen) first case. Choosing easy assignments with minimum calculations enabled the three amigos to polish off their first cases while the boss was still occupied with her admin – allowing them to cherry-pick their second cases too. They could then let their boss allocate their final cases from the difficult pile, secure in the knowledge that they could finish even the toughest ones in the half of their shift remaining.

By completing a ‘hard’ case every night, they’d be seen not only as diligent and efficient, but also as ‘team players’. Best of all, though, their strategy ensured they always hit the three-a-night target.

So where’s the harm, you may be thinking – a little gaming of the system, but the required work is getting done, isn’t it?  For sure, three cases are being performed. But productivity is being manipulated, so the figures are unreliable. Cherry-picking may mean that some cases are ignored altogether, while the relationship between manager and workers seems to have taken on a latent parent-child dimension.

More worrying is that I’ve observed similar behaviour – cherry-picking and other means of gaming the productivity numbers – in almost every organisation that employs activity targets (that is, targets based on measures of efficiency – service levels, number of rings to pick up a phone, hours or days to respond to a call – as opposed to overall effectiveness). The exceptions have displayed related but distinct dysfunctions.

How to meet targets 2

For example, in the bereavement-handling back office of a large retail bank – where the accounts and financial holdings of deceased customers are closed down or otherwise resolved at the request of surviving relatives – the staff were again working to an activity target: in this instance three ‘cases’ an hour. As in the complaints-handling example, the workers had worked out ways of achieving the target even though, in this case, the IT system prevented cherry-picking.  Instead, workers would send out letters requesting information already on file or clarifying points already understood, fulfilling their target but not the relatives’ wishes, and building a backlog of unfinished work in the department.

The frontline workers weren’t necessarily being insensitive. Under pressure of the 20 minutes a case, it was easy to miss that their branch-based colleagues had noted these details upstream. Newly hired staff on probation were under additional pressures, since failure to meet their target by the end of their induction meant they would not obtain permanent contracts.

In haste, volume targets were often met at the expense of quality. Most of the mistakes were small, but in the circumstances even minor errors such as continuing to send out a mortgage statement in joint names could be deeply upsetting – and customers often phoned in to say so.

Treating quality as a separate issue to the enforced productivity, different teams specified procedures for ‘error-proofing’ common mistakes and then inspected both the work of the workers and that of the checkers. The wasted effort was significant, while error rates remained unchanged over time as their underlying cause – haste – was ‘designed in’. In one of the most ironic features of the whole setup, quality and error rates were themselves subject to targets, with predictably perverse consequences.

The worker or the system?

The problem in these examples is assuming that the workers can be held exclusively accountable for their work. That is – in the complaints-handling example – assuming that a colleague completing fewer than three cases a night is incompetent or ‘shirking’, while a colleague who chalks up better than three is a ‘hero’. In this way performance is personalised, while many other explanations for differences in the quantity of work produced by individuals is ignored – for example, clarity of instructions, complexity of the complaint, number and nature of the product holdings, nature of the calculations for redress payments (where applicable), degree to which IT supports the work and the ease of accessing and using it, presence or absence of activity targets and degree of knowledge and skill of the worker (resulting in turn from selection and training processes of the organisation), and the motivation of the individual (linked as it may be to the hours of employment, nature of the work, environment and a myriad of personal factors).

The same assumption – that workers are the dominant factor in performance – is at the heart of forced-ranking (‘rank-and-yank’) performance-management systems, in which ‘lowest’ performing colleagues over a period are ‘managed out of the business’, in other words, fired.

Yet we know from the writings of W. Edwards Deming among other authors that performance variation between individuals in any team is governed by the system in which they work. In simple terms, people need three things to do a good job: information about what to do and how well they are doing it, equipment that is fit for purpose and simple to use, motivation undistorted by external incentives, and an interesting job to do. Where these elements are in place, the creativity of colleagues and managers can be brought to bear on improving work, rather than gaming it.

Don’t game the work: improve it

In such an environment our three amigos and their boss might jointly be working on improving the reliability and consistency of the redress calculations or the decision-making of colleagues, or eliminating the cause of the complaints in the first place. In the bereavement-centre example, colleagues and managers could work together to resolve cases in the most timely fashion, anticipate predictable follow-on requests from solicitors for tax deduction and interest payment certificates, and improve access of colleagues to ‘specialist’ product information that is not ordinarily accessible via the IT platform.

Without consideration of the system in which people work, ‘performance’ will likely be attributed to them. Even though they are well aware of many of the causes of performance variation, the primacy of activity measures – reflecting an underlying preoccupation with cost – and the rigidity of performance-management practices – reflecting a belief in the primacy of personal determinants of performance – mean that those individuals will do whatever it takes to survive in the system of work in which they find themselves, even if it means playing dice with productivity and the quality of the work.

All of which is to say that I didn’t get much work done on my way to Manchester. But I did have an interesting journey.

---

will.pyke@vanguardconsult.co.uk

Read similar articles in Edition One of The Vanguard Periodical: The Vanguard Method in Financial Services. Download it here