I spent most of my career in large companies before setting up a software company, and I now run a small analyst firm. One of the things that has intrigued me about the difference between large and small companies is the difficulty in getting staff to act in the interests of the business rather than that of their own little corner of it.

To illustrate what I mean, take a couple of examples from my days at Shell. My first project there was reviewing the IT strategy of a refinery. I was bewildered to discover that they had recently built their own datacentre, when one of the largest datacentres in the group was just a few miles up the road. Surely the economies of scale would mean it was more efficient to buy services from the large datacentre? “Ah, but we can do it cheaper” was the response. How could this be, given everyone was on the same pay scales? When I looked at the figures, the only reason why it seemed cheaper was that the refinery did not bother to pass on a slew of overheads to its IT group, so essentially everything it did was subsidised. Strip this out and, guess what, it was financial nonsense to have a separate small datacentre.

The people who made this decision were not stupid, but at that time having more staff and budget meant power and possibly a higher pay grade due to all that extra responsibility, which may not have gone unnoticed by the manager who recommended the datacentre, even if he knew that the business case was dubious.

A more extreme example was a poorly-set sales commission scheme in South Korea, where salesmen were rewarded on revenue with no account of the profitability of the deals being made. Two of them did exactly what the scheme rewarded them for, and started giving huge discounts, racking up large increases in volume sold but decimating the profit performance of the business unit.

Lest you think this kind of thing was peculiar to Shell, I recently spoke to an ex-colleague who now does high-end IT consulting. At a recent assignment at a large multinational he was asked by the CIO to look at how to cut IT costs. He invited himself to project steering groups where he would ask: “Who is the business customer for this project?”. Often there was much shuffling of feet, but he would insist on getting the name of the person supposedly sponsoring the project. He then simply rang up the sponsor and asked whether they had ever heard of the project. Frequently they had not, at which point he cancelled the project. In a year he saved $100m of IT expenditure, in large measure by checking whether the large projects the IT department busied itself with were actually needed, then ruthlessly purging any that were not.

The point is that the staff in question no doubt felt they were doing a decent job, yet the system in which they operated allowed them to squander vast sums of shareholder money. Perhaps they had an inkling that not all was right, but the incentive systems and politics of the firms did not cause them to raise a red flag; indeed, perhaps they were discouraged from doing so by their supervisors.

Each of the companies in question had staff share schemes so, in theory, staff interests were aligned with those of the company. However, as Lehman Brothers’ demise demonstrates, this is not enough (Lehman’s had a very high 30 per cent employee share ownership). Using balanced scorecards when setting staff remuneration? Perhaps these can help, but they are fiendishly hard to design in a way that is both understandable and does not enable staff to play the system.

It is remarkably hard to get people to act as if spending the company’s money was really their own. The notion of ‘clawback’ compensation, where bonuses are tucked away for years and potentially reclaimed by the company if something later comes to light, seem to me very difficult to design fairly; staff may simply ignore them as a method of motivation if they suspect that the company can just change the rules later and not pay them.

There is no magic bullet, but people who feel proud of their company will generally try and do the right thing, whereas if they believe it is ‘just a job’ they may well be tempted to act in their immediate personal interests rather than those of the corporation. Hence, it is important to measure just how engaged employees feel with the company via surveys, and treat this measure seriously: a slump in how engaged employees are may be the first sign of serious trouble ahead. Staff on the ground will often be the first to know of poor practices, yet frequently just go along with the herd. Creating an environment where it is OK for staff to challenge apparently poor decisions by management is the first step to heading off some of the bad decisions that companies make.

More Andy Hayler articles:

Andy Hayler: A bit of discipline required

Andy Hayler: Bigger is not always better