Of course, it’s important to track progress and monitor performance, but the way this is done in many organisations can be counter-productive.
There’s too much focus on short-term “instant metrics” and too little focus on the connections between different metrics or whether those metrics serve any purpose at all.
It’s like “instant food”, such as Pot Noodles You’re hungry. You get a quick dose of energy…instant feedback from your gut telling you to feel better
Once in a blue moon, eating a Pot Noodle is unlikely to do any harm. But eat Pot Noodles for dinner every day and pretty soon you’ll be fat, bloated, and out of control.
Just like the management and administration overheads in organisations who have hooked themselves on the dopamine hit of “instant metrics”.
The overhead involved in tracking and reporting all these stats, and the management infrastructure required to spend all day in meetings talking about this tsunami of instant metrics is enormous. And the upside is often somewhere on the spectrum between small and non-existent.
But, whatever those metrics might be tracking, the root cause of their ineffectiveness is often the same. For pointing that out, we have British economist Charles Goodhart to thank.
Goodhart’s Law
Although it’s not quite how he originally expressed the idea, in the way it’s usually written today, Goodhart’s Law states: “When a measure becomes a target, it ceases to be a good measure.”
There are several potential reasons for that, but first and foremost is that an organisation is likely to distort what it does to ensure the delivery of their key targets, even if the end result disadvantages the organisation. But nobody cares because “all the targets have been met”.
In my professional life, whenever I’ve seen a business in real trouble, it’s rare than they have anything other than a stellar monthly KPI report or balanced scorecard.
There might be a couple of amber-rated measures on there to give a superficial level of reassurance to the casual observer, indicating there has been some rigour in the assessment process, and that people are being held to account for their performance.
But the list is mostly rated green, or “on target”. And red ratings are incredibly rare.
Results not tasks
At heart, those measures have often been selected because they are largely within the control of the organisation, even though they mean very little in terms of bottom line performance.
For example, a marketing team might have a lead generation target. That’s a task. An internal activity that, subject to making the budget available and employing reasonably competent people, can be delivered relatively reliably.
However your business earns nothing from a lead. You don’t get paid until someone makes it into a sale.
The “generating a lead” bit is a task. The “making a sale” bit is a result.
In general, organisations build revenues, profits and cash flows by concentrating on results, not on tasks.
Sure, you might want to track tasks to get some idea of the future trajectory of the business, perhaps to manage your productive capacity effectively, or decide whether or not you want to approve a member of staff’s holiday request.
There is an undoubted benefit to tracking metrics like this. To use Goodhart’s language, that makes tasks a measure, not a target.
The only targets should be the results, because it’s easy to game tasks, but it’s much harder to game results.
If my target is to get more leads, I can go and buy some terrible leads very cheaply, turn my monthly KPI sheet green and apply for the next promotion that comes along.
If my target is to get more sales, I either get them or I don’t. Unless I’m thinking of committing fraud (spoiler alert: I wouldn’t) the only way I get to book a sale is if a potential client signs a contract and we deliver whatever it was we promised them. That’s much harder to fake.
Harder…but it’s not impossible
It’s a lot harder to fake a sale than a lead, but it’s not impossible.
Given the choice between those two metrics, I know which one I’d be choosing. It’s the sale every time for me. The end of the process, not the start of it. The result, not the task.
But even there lies a potential problem.
What about if I’m selling to people who aren’t going to pay us? Or who are going to complain and demand credits to an unreasonable level? Or who are going to sue us for some problem that was caused by their misuse of our product, rather than a failure of the product itself?
Using results (eg sales) to assess performance is vastly better than using tasks (leads). But what we really want to measure is cash left in the bank at the end of our guarantee period, so we know that the customer has paid in full and we have no continuing liability for the product we supplied.
For most organisations, this is impractical. The elapse of time between making a sale and the end of any guarantee period is too long, they believe, so all the internal performance metrics are based on tasks, not results.
Measuring tasks
I get the attraction of measuring tasks.
It’s like chugging back an espresso every half-hour or so. The caffeine hits keep coming. People get addicted to them and lose track of the organisation’s real objective – delivering results.
In fairness, it’s also much easier to measure tasks. Someone clicking a link to download a lead magnet is childishly simple to measure. Just about every piece of marketing software can spit out an hourly/daily/weekly report showing exactly how many people have clicked the link, and who they were.
No judgement is required. The lead is in your CRM or it isn’t. It’s a yes/no answer.
But there isn’t enough nuance.
Who downloaded that lead magnet? A genuine sales prospect…a “tyre-kicker” who will take up a lot of time and trouble, but ultimately not buy anything…someone doing research for a college project…a competitor trying to clone your sales process…?
The list goes on and on, but the point is this – at the task level you have no idea whether someone clicking a link is good (potential customer), bad (competitor trying to rip off your sales process), or neutral (student doing a project).
This is why, the eventual failure of failing organisations is always a huge surprise to the people running them, because they’ve been fooling themselves for quite some time that the “all green” RAG rating means they’ve been doing a great job.
Usually, by carefully choosing the tasks to measure to make sure they have a high degree of control over their delivery, meeting targets isn’t all that difficult in those organisations.
The bit they often miss is that they are focusing on the wrong targets.
Measuring results
Measuring results is harder than measuring tasks. They tend to be more multi-dimensional because anything worth achieving is almost certain to have lots of moving parts.
That multi-dimensionality mitigates against simplicity in reporting because very few things are necessarily good or bad. Lots of the individual components are in the “it depends” category of decision-making, where judgement is required and a spreadsheet or an AI bot won’t be able to give you a definitive answer.
Imagine you’re approached by a new customer who wants to buy from you, but at 10% below your normal prices. Should you take that work?
Well, it depends…
If it was for a large order, I might be prepared to give some sort of volume discount because I’ll get some economies of scale in the factory.
But equally, the factory might be quite busy at the moment and a large order at way below our normal prices might not leave enough capacity for better-paying customers.
On the other hand, if it was the holiday season and the factory was pretty quiet, I’d probably take the work so that we were at least making a contribution towards our overhead costs.
And let’s not forget this might be our first contact with a hugely prestigious client who had just been let down by their normal supplier. Might we want to pull out all the stops for them this time round to get our foot in the door with a perfect new customer?
For all those reasons and more, results are not nearly as easy to fit into a neat KPI dashboard as tasks are.
So they are rarely tracked. Rarely mentioned. Rarely the subject of discussion at a board meeting.
Whereas tasks tend to get picked over in minute detail if, heaven forbid, one of those KPI metrics starts flirting with amber territory on the dashboard.
Check for “Goodharts” in your organisation
In our data-driven world, it’s tempting to think all data is good data. But that’s not the case at all.
Data measures which can be fiddled – as a lot of task measures can be, pretty easily – are of very limited use to an organisation. They are interesting as far as they go, that’s all.
The fact that task measures are easy to track and report is not a good reason to stuff your KPI dashboard with any more of them than you already have.
As Peter Drucker said, “There is nothing quite so useless as doing efficiently that which should not be done at all.”
So take a look at your KPI reports, your balanced scorecards, and whatever goes to your board on a monthly basis for performance tracking.
The likelihood is that those reports mostly track tasks, and barely mention results. And if they do, it’s also likely that Goodhart’s Law applies – the metrics you track might have morphed into targets. Tasks which are carried out irrespective of their impact on your bottom line.
It’s not hard finding examples of Goodhart’s Law inside most organisations. So, if that’s what you find, it might be time for a rethink.
After all, tasks don’t pay an organisation’s bills. Only results do that.
Just in case it’s bugging you, the title of this piece is a poor attempt at humour, based on the lyrics of Feargal Sharkey’s 1985 UK Number One record, “A Good Heart”.
The original lyric was “And a good heart these days is hard to find”.
This article first appeared on LinkedIn.