The last thing your business needs is more metrics

On the basis that no less an authority than "In Search Of Excellence" author and management guru Tom Peters was kind enough to tweet this article from my own website, I thought it was worth sharing here too in case you missed it first time round...

The problem with metrics is that smart people try to game them, and dumb people try to achieve them.

You know the sort of person I mean…the one who dogmatically insists on continuing to do something they get measured on, all the while turning a deaf ear to the legions of customer complaints, wielding their weapon of choice… “company policy”… as the excuse for disappointing paying customers..

Difficult though this dynamic is for any business, the problem is more acute the more metrics, targets and KPIs your business has. I once ran the division of a multinational business where we had 42 different metrics to track and report on every month.

Each of these numbers had some degree of relevance…they were interesting, up to a point…but with 42 metrics to keep on track, that gave me about an hour a week on each of these apparently equally-important metrics.

Not that I actually had an hour a week to spend on any of those things. My experience is that businesses with lots of metrics also tend to have lots of meetings where the metrics they’re tracking are discussed, analysed, picked over and post-mortem’d.

Every metric had a manager somewhere who was responsible for that particular measure across all the divisions of the business, who in turn reported in to someone pretty important in the Head Office management structure.

Because each metric’s manager reported to a senior person at head office, each of them insisted on their own battery of meetings and reports…separate from, and in addition to, the corporate set of meetings and reports through the existing reporting structure…so each metric’s manager knew what was going on well enough to pass on their own reports through to their own bosses..

If you think that sounds like chaos, you’d be right. Very little ever got done to move the business forward. All our time was spent reporting on history or having meetings about it.

Here’s a new initiative – let’s make sure we’ve got some metrics in place

Part of the problem in that business was every time a new idea was sanctioned by the CEO or the Board, a raft of new metrics was brought in to make sure this new idea could be measured and tracked in minute detail across all the operating divisions of the business.

Thankfully, as you might have realised by now, it wasn’t very often that a new idea was actually implemented. That would have required us to stop talking about things that had happened in the past for long enough to have an idea or two about what we might like to do in the future.

As a Finance Director and CFO, this approach used to drive me crazy, (although I wasn’t the FD or CFO in this particular business, so just had to go along with it).

The amount of unproductive time spent in senior managers meeting with one another was beyond astounding. I reckoned the costs of measurement and reporting was possibly the single biggest cost the business had, even though resources were being starved from the front line at the time on the grounds of “cost pressures in the business”.

For every new XYZ Initiative Manager appointed, we could probably have put four people on the front-line where they were desperately needed. But the Chair, Board and Chief Executive of this business could never see things that way.

I’ve no doubt they were well-meaning. But they’d all taken on board the sort of thing people are taught at business school somewhat too literally, which made the whole thing unmanageable.

As a principle “what gets measured, gets managed” has something going for it, but carried to the extreme, all it does is seize up the inner workings of a business.

Of course businesses should have some basis for deciding whether the initiatives they’re implementing have some positive benefit to the business and its customers. But there are often simpler ways of achieving that objective.

After all, when you break it down, all any business is trying to do is bring in enough income from customers to cover its costs and leave a profit.

Since profit is calculated as a balancing figure, you’re only really managing two dimensions here – income and cost. Profit is what’s left over after deducting one from the other.

In most businesses, costs are either fixed (that is they don’t wary much from month to month, like salaries or rent) or variable on some predetermined formula (X per hour, Y per item produced, Z% share of income generated) or a mix of both, again on some predetermined basis (eg some permanent staff paid on a salary and some temporary staff paid on an hourly rate)..

Off-hand, it seems to me that the optimal number of metrics and KPIs to measure those mathematical relationships is some number a long way short of 42.

You might say…”that’s all well and good, but what about non-financial measures that matter to stakeholders?”

Fair point – how many of those do we need?

Treat people fairly, without bias, based on gender, ethnic background or belief systems? I’m all for that…

Ensure your people are safe from harm at work? Seems reasonable…

Buy your products fairly, without supporting exploitative labour practices in third world countries? Yup…

Reduce our impact on the planet? Another tick…

Operate according to the principles of good corporate governance? Yes to that one too.

There may be some industry-specific concerns over and above those. If you’re disposing of radioactive waste from a nuclear power station, I’d expect a few additional metrics around how you might do that without destroying half of the nearest town, for example.

But those should be pretty rare. There can’t be many businesses which need to track more than six or seven key metrics in detail. And even if you could, that doesn’t mean you should…especially when you consider the cost of collecting all the data you need for a never-ending round of review meetings about the latest metrics or KPIs.

The cost of measurement

You’d probably come to the same conclusion if you did what very few businesses bother to figure out – the cost of measuring and reporting on each initiative you think is important.

In the business I described earlier, pretty much every measure had its own manager to track each metric, so there was probably a 60-80 grand salary just in that.

But it gets worse…

Some of these people had their own departments to help them with their analysis and reporting. One, I recall, had a department of six, in addition to the manager, all on 40-50 grand salaries.

Measuring this particular feature of the business probably cost the thick end of ￡400,000 just in salary costs…maybe as much as half-a-million when you add in IT costs, facilities, travel and the costs of the conferences these people always seemed to be going to.

And that’s just the more obvious measurement and management costs.

On top of that there was my time, time spent by the other divisional heads, time spent by our respective teams collating the information in the first place…which, in some cases, required an investment in bits of kit or software where human beings just couldn’t collect the data quickly enough or accurately enough..

I couldn’t give a definitive picture of what that cost might add up to, but I guesstimated it at about half again of the costs incurred running the official “measuring departments”. And I’m probably being kind at that.

Although the 6-person measuring department was the worst example, multiply that guesstimated cost by 42…one for each metric which was “vital” to the business…and add on half again for our “in-house” costs at divisional level. You’ll quickly see why the cost of measurement was easily one of the biggest costs in this business, even though nobody ever thought of it like that..

It’s been a while since I worked at this business. They may have changed since…although I wouldn’t bet on it. Some behaviours become too deeply ingrained to be changed under anything other than the most severe existential threats.

Don’t measure it, do it!

That business was the first thing going through my mind at the weekend as I read a succession of articles about changes people felt should be made to the business world.

These were changes for the good, let me add.

The articles suggested encouraging more women and minorities into senior roles, emitting less carbon from company vehicles, providing opportunities for young people trying to get onto the job ladder and so on.

I wouldn’t argue against any of those things.

However every article demanded additional metrics, tracking, targets and KPIs against each of those inherently worthy objectives.

But, thinking back to the place I used to work, I imagined the impact on that business if they picked up any one of those and started to carve out a six-person department to track progress and report to head office about it.

What would you rather have – ￡500,000 spent on measuring the recruitment of women and ethnic minority candidates into senor level posts or ￡500,000 spent giving women and ethnic minorities within the business the training, support or coaching they might need to be a credible candidate for the next senior role?

Which one is more likely to create the positive result you’re looking for…measuring it or doing it?

To quote one of the world’s foremost quality and efficiency experts, W Edwards Deming (“Out of the Crisis”, p15)…

"…measurement of productivity does not improve productivity."

I’d rather spend the money actually doing things that lead directly and tangibly to the outcome being sought, instead of spending the same sum measuring whether we’ve done it or not.

In fairness, I don’t think this is a position most people would argue with. The problem is that very few businesses know the true cost of their back office activities…including things like measuring and reporting on information, and the costs of managing those who do.

And, admittedly, while it’s not impossible to work out, it is tricky…and sometimes fairly approximate unless you have everyone in the business fill in time sheets all the time (which would, of course, require a Time Sheet Manager and a raft of people to collate, do the data entry and report on whatever the time sheet data told them…rather defeating the point!).

There are times when the cost of getting more accurate measurements just isn’t worth the cost involved. Intelligent approximations, consistently applied, will often get you close enough for most decision-making purposes.

When is the cost of measurement too high?

It’s hard to be dogmatic about this. If you’re servicing aircraft engines or conducting open heart surgery, you might think it was important to do a bit more tracking than you might for a business making black plastic refuse sacks.

But my personal “concern meter” goes off when the costs of administration are more than 5-10% of the project concerned.

On that basis, a budget of, say, ￡500,000 should have no more than ￡50,000 spent administering it and reporting on it. The rest should be spent doing whatever the objective might be.

In a large corporate environment, ￡50,000 doesn’t get you very far. It’s a small fraction of a C-Suite executive’s time. It might get you an Executive Assistant, or half a departmental manager…maybe most of a front-line manager.

When you approach measurement and metric setting from that perspective, it forces you to think about what you really need by way of metrics, KPIs, dashboards and whatever else you use in your business.

When you’ve only got a few grand to do your measuring with, you’re forced to concentrate only on the metrics that really matter…not every metric a full-time department of six might come up with between them after an all-day brainstorming session..

And in case you think this is impossible, I assure you it isn’t.

I once took a major organisation through an environmental certification process, which was important to them at that point in time for reasons I won’t go into here. I did it with a budget of zero and just little scraps of people’s time.

We invested every penny we could leverage out of existing budgets to support the process as best we were able and, in just a couple of years, went from being one of the sector’s poorer environmental performers to gaining a prestigious certification from an independent awarding body, which, in turn, helped boost our market positioning and ultimately our revenues.

We only had one metric – achieving, and subsequently sustaining, that accreditation.

I led meetings about once every two months, for a couple of hours each time, so we could catch up with the dozen or so people across the business who might be able to impact on this initiative.

At that meeting we identified what cash had been freed up from existing budgets and prioritised where we would spend it to make the fastest progress possible towards one or more of the certifying body’s assessment criteria.

In the three years or so I ran this initiative, I made one report to the Board, about 18 months in, to explain what we were doing and how we’d spent the money we’d found in pursuit of the objective to gain that certification. We did another session when we’d achieved the accreditation. That was it – maybe an hour in total over three years.

That apart, I refused to set up meeting cycles, reporting sessions, KPIs to track what people were getting up to or any of the usual stuff that people do in this situation.

There was no need for any of those things as I could manage it all through the bi-monthly meetings. So we spent little or nothing on administration, measurement and management. We spent everything we could achieving the objective.

Now I’ve got my good points, but there was nothing I did that nobody else couldn’t do with a bit of trust and some room for manoeuvre. There was no financial risk as everything we spent still had to be signed off through the normal approval processes.

My experience is that, given the will, and given an understanding of just how much management and monitoring costs the average business, there are very few projects which can’t be delivered the same way. You might surprised how quickly you can make progress, if anything.

But we can’t trust our people to operate like that…

When I tell this story to business leaders, I’m often told that they could never do things the same way because they can’t trust their people enough to operate without lots of reports and meetings.

I find this very sad…but this perspective raises an interesting question…

Who hired these people in the first place?

Well, it seems that was the same business leaders who claim they can’t trust the people they hired themselves!

Seems to me the solution to that problem is entirely in their own hands. Hiring people you can’t trust doesn’t sound like the smartest recruitment decision I ever heard.

“Oh, but the people we could trust are too expensive,” people sometimes respond, “so we only have the budget for entry-level people.”

Again, the solution is in the hands of leaders. Running the business on cheap labour is often the most expensive way to run a business.

On the surface it looks cheap, but by the time you add in management, reporting, tracking and measurement systems sufficient to give at least a reasonable level of assurance that nothing is going badly wrong, this approach is often the most expensive option.

But this response is also an illustration of why the measurement, tracking, reporting process is fundamentally flawed.

Despite all the measurement, tracking and reporting they already have to regulate the behaviour of people they employ, those business leaders still don’t trust their staff.

More metrics isn’t the answer. They’ve got plenty of those already. But none of them do the job they really need doing.

Metrics won’t give you trust. At best they protect your downside a little, but they’re not infallible – many a formerly market-leading business has gone into administration even though all their metrics and reports were pointing in the right direction up till the moment the bank pulled the plug (Patisserie Valerie, anyone?).

Trust is an emotional construct, not something governed by the laws of physics which you can measure mechanically and improve in a factory.

If trust is what you want to create, don’t imagine traditional reporting mechanisms will help. They won’t.

Instead, work on trust (pausing to reflect, at least briefly, on whether the problem might be with you, rather than with your people).

Get that in place, and redeploy most of the resources (keeping back a reasonable 5-10% so you can do the tracking that really needs doing) spent on managing, reporting and tracking things into making your overarching objective happen.

From the moment you do that, your business will march towards your objectives faster than you can imagine. You’ll get where you want to go faster, more profitably and with a lot less stress and strain on your life and your calendar.

And if you think this is just my own personal opinion, no less an authority than Dr W Edwards Deming, the father of modern quality control, had as part of his famous 14 Principles “Eliminate management by numbers and numerical goals, Substitute leadership.”

Go on. Give it a go. You might be surprised how good your business can be when you don’t measure things.

This article was first published at FinanceDirectorCFO.com on 18th June 2019.