Yesterday I found myself in a couple of conversations that revolved around what has been a bit of a recurring theme for me – the dangerous downside of measuring things.
The first was around the gender inequality evident in the IT industry. My own view is that this is a symptom of a more pervasive lack of diversity of thinking within the industry, that inevitably leads to poorer technology being delivered overall (because it represents the output of what is a small meme-pool). Gender splits of employment are a barometer of this issue, probably one easier to measure than the unknowable of “things that would be better if IT comprised of more diversity of people and thinking”. But making gender splits of employment of IT the goal of change initiatives misses the underlying causes, and would be akin to running mass campaigns to attract racial diversity into the police force without addressing underlying institutional racism.
The second was a discussion about The Telegraph’s new strategy and how writing for search is seen as a core skill journalists now require. Again, the consequences of judging journalistic success by search ranking will be news that is designed to rate highly in search results; the intent will almost certainly be lost.
The two laws that I often cite on this topic – Goodhart’s Law and Campbell’s Law – explore in more detail why number-based goals. The generation of managers who have been indoctrinated with “SMART” objectives haven’t helped matters either, as many misinterpreted “Measurable” with “a measure”.
But put simply, if you want to get people to focus on getting something done, make sure your goal is clearly stated in verbs and nouns, not numbers: “We want to make IT a more accepting and diverse industry where gender, ethnicity, sexual orientation or disability aren’t barriers to us delivering great services to our clients and customers”, for example, or “We want to create news content that informs and delights readers, attracts advertisers, and is fit for the evolving world of digital.”
Numbers might be used to track progress towards those goals, but if they become the primary focus they will change in their meaning because of that new purpose, and people will cheat to hit them anyway. Unfortunately in our now data-obsessed evidence-driven world, it seems that challenging the primacy of numbers is the modern day iconoclasm.