Imagine a group of 50 people with some connection between them, maybe a neighbourhood, or a company, or simply a loose set of connections. We take the body temperature of each of them, and we calculate the average. It’s 36.3 degrees. And we declare that to be normal. Because it is, according to what has conventionally been agreed. The group temperature is 36.3°.

About 15% of the group is composed of very ill people, but their illness does not manifest itself in changes in body temperature. Their thermometers don’t say anything about that. One can be very ill with a normal temperature, and also marginally ill, with a common cold for example, and have a serious temperature. The ‘average temperature’ does not tell us much. Certainly, with the thermometer data we have no idea what will happen to the very ill. ‘The average’ translated into a number does not have a predictive value.

Six months later we repeat the exercise and, for whatever reason, the average temperature is now 36.9°. This is still normal in medical terms. Something is going on here. Maybe. But we don’t know what to make of it. Over those 6 months, a few of the very ill died, some who were normal became ill and a small group had been infected with Covid. But we are not looking at any of these things. The average temperature is still OK. We are in charge of temperatures, and we report on temperatures. We can also report an increase of 0.3 degrees over a period of 6 months.

We’ve increased our knowledge of ‘thermometer management’ a bit, and we have a bigger set of temperatures, a larger database. We can say this confidently. Overall, everything remains ‘normal’.

The absurdity of this thinking may be too caricature-like to take it seriously but it’s not miles away from what people are still doing in the area of Employee Engagement, which still dominates the airtime in some corporations, even if it is with less intensity these days. We look at scores, compare scores and celebrate or deprecate its ups and downs. It’s, by and large, score management.

The need to reduce data to simple, aggregated parameters (‘scores’) is understandable but it forces us to artificially simplify things under a scientific illusion. This illusion of control is significant in management thinking. It provides both comfort and a pseudo-legitimization of ‘data’, particularly for people who profess an allegiance to anything ‘data driven’ or ‘evidence based’.

Aggregation of bits and pieces under a label (‘engagement’) may be neat in the above context but, when practiced by ‘data driven’ and ‘evidence based’ warriors, makes it a bit ironic.

Many years ago, circulating in the media, there was a picture of a signpost at the entrance of a town in the USA. I never figured out if it was real. It looked real. The signpost said: ‘Welcome to Centerville, population 650, churches 5, schools 4, saloons 7, total 666’.
I would say ‘their score was 666’.

The ups and downs of employee engagement scores have been the ritual game in many corporations. Less so today. Everybody agrees that reducing the conversation to what happens to ‘the score’ is a sad state of affairs. Many of its defenders would argue that what matters is how the score was achieved and that the system provides well for an analysis of this. I agree. It usually does. The problem is that no matter how well intentioned this is, the airtime tends to be dominated by ‘the score’ and its yo-yo properties. Over the years, many executives have told me in private that the ‘score cult’ is a bit of nonsense. In public, however, everybody has been more prudent and not challenged it.

This is not a rant about measuring. Far from it. Measurement is important, but its reduction to semi-absurd proportions needs to be called out. Engagement, as in employee engagement, is always associated with ‘scores’. It is, amongst other things a military term. This is where ‘rules of engagement’ comes from. We have adopted the term gladly and blindly and now we talk about engaged or disengaged employees. When you look at the use of the term, we can find all sorts of overlapping meanings that come and go depending on the conversation. For example, employee satisfaction, job satisfaction, and also, motivation, sentiment, even morale, and, of course, ‘culture’.

The productivity magnet

For me, the most damaging use of the employee engagement framework is the widespread one that states that ‘engagement’ predicts productivity. A poorly engaged workforce (whatever that may mean) will have (pick a score), a 15 or 20 percent, or more, decrease in productivity. So, boosting engagement will boost productivity. And this is good. If productivity is what you are aiming for.

This model reduces human interaction and the collective action of a company, for example, to a machine with inputs and outputs. More engagement, or more satisfaction here, is not good in itself. It’s only good in so far as it produces more. Years ago, I was struck by the apparent seriousness of a 2012 business book entitled: ‘Contented Cows Give Better Milk: The Plain Truth about Employee Relations and Your Bottom Line’. Unchallenged, and also praised as a bit of fun, it was telling the truth about many corporate approaches and a great deal of the HR and consulting industry. We want a more engaged workforce (pick a definition) because we want more productivity, more milk. Not because it may be good in itself, because we want people to be fulfilled by the job, because we wish that satisfaction to be part of what working here is about. No, It’s the milk stupid. The fact that not many people have shouted, loudly, about this insulting view of human nature, tells us that in the area of management of organizations, we can still get away with murder.

I am not against scores or indicators per se. I understand that, particularly in a large organization, that may be the only way to put some order to complex data. I am against a single indicator when it comes to ‘explaining culture’. The single indicator, single joystick, is absolute nonsense. As I said before, many people may push back and say that this is not how it works, that they always take into account other data. Great. I hope so. But this is not always my experience, having been part of many conversations about the single ‘employee engagement score’ and how good or how dreadful it was or is.

As per other ‘indicators’ or pieces of information that could give context to ‘the score’ (in case you are still married to it) there are many. The ones that have been used successfully are in areas such as people turnover in a given period of time, qualitative data from well structured, third party managed, exit interviews and my own favourite ‘Stay Interviews’ (Why are you still here? Even better, why on earth are you still here? Literally…). Of course, some of these require more time and resources but, hey, I thought that the employee was the most important ‘asset’.

It’s surprising how, in many areas of the company business, we are used to using a set of parameters to define what is going on. Commercially, we would never assess company function by one single type of data, for example, sales. Sales means a lot for those in charge of Sales. But commercially you would also look at the market share, earnings per share, profitability, margins and so on. Equally, it’s a bit pointless to focus on employee survey results only, out of context and in isolation from a wide set of ‘cultural parameters’. The problem is not that any sensible leader would agree with this, but that the same sensible leaders may spend a lot of time ‘on the scores’ just after agreeing that it does not make a lot of sense.

The thermometer approach that I described at the beginning is incredibly inaccurate, but incredibly simple! It banks on calendar points that also tend to be well known: ‘hey, employee engagement survey comes next week’. In other words, here is your public invitation to shout, complain, rant, or simply rate things. It’s not a week like any other week. It’s the Survey!

This model is broken. Business organizations should adopt well established practices from ‘alien territories’ such as political campaigns. In particular, what I call ‘continuous sampling’. You are always sampling here (on whatever you want to ask) but never to everybody at the same time and, as much as possible, always silently: this neighbourhood, that group, that sector, 10% of people in X, 15% from Y, etc. You are always gathering data in the background and then making some sense of it, a bit at a time. Focus groups have had their dose of ridicule in popular media but, call it what you wish, there is no alternative to talking to people. Of course, we have to deal with confirmation bias and many others but, you would have thought we would employ professionals able to deal with them.

Today there are also some technological approaches such as ‘pulse data’ from apps that allow for real time insights. All of which have the same caveat: don’t use one single criteria but as many data points as possible from different places and times, and then use your judgement. Sure, engagement is whatever you have decided it to be. For example, a mixture of satisfaction, an interest in the company and feeling like a participant and an actor (or not) and a view on management. But ‘the total’ cannot be 666.

There is a ‘higher level’ problem here which is usually overlooked: the hidden assumption (and perhaps desire) of a single concept of ‘good corporate citizenship’ which means, by default, well ‘engaged’.

Here is Anna. She does her job promptly and proficiently but is not interested in anything else. Her primary goal is to earn some money, pay the mortgage and save for a gap year. Anna does not click with any talking on missions and visions and when she attends the Town Hall it is mainly because most of her mates do. It is a break. They also show pretty videos there.

Meet Peter. He is a rather frustrated manager who thinks that people are mostly taking him for a ride. He feels he needs to play ‘strong man’, and he does. He is very clear on what he would like to change in the company so that people are more productive. He is often involved in this kind of conversation when talking to colleagues.

In those scenarios, is Anna disengaged and Peter engaged? Is there anything wrong with Anna? With Peter?

Many moons ago I worked for an American boss who repeatedly told us that he wanted everybody coming to work in the morning with a clear view of the mission and vision of the company, and fully committed to them. For him this was ‘the car park test’, as he put it. Were we really working towards everybody having that feeling in the carpark in the morning?

This naïve concept on ‘engagement with the company’ or, for that matter, ‘alignment with the values and goals’ tends to deliver things that mostly pay lip service.

Company culture

Company culture has suffered a similar set of problems when a reductionistic approach is used: culture equals employee engagement and employee engagement equals the score in the employee survey (because we have thermometers). Here is a bit of a circular definition which is real (I am not making it up): definition of Engagement is whatever you have decided Engagement is and you have decided to measure via an Employee Engagement Survey. In summary, Employee Engagement is whatever the Survey says. And it says it with a number, which is very handy. It’s time to abandon this.

It reminds me of my student days in Psychiatry when we were studying Intelligence. We used to say, comforted by a smile that pretended to project cleverness, that Intelligence was whatever was measured by Tests of Intelligence. That sorted out a lot of conceptual problems!

Culture is a big envelope. Never mind the dozens of official definitions, people engagement is one element. Assuming you are going to define ‘engagement’ and ‘culture’, the connections between both will be the consequences of that definition.

For me, company culture needs to be understood as a Petri Dish, ‘where things grow’ (the latter, my definition of culture, surely influenced by my Life Sciences background). What we put in the Petri Dish will grow and produce something. This ‘something’ will be labelled in multiple ways, because, amongst other things, we love labels: empowerment, accountability, trust, and many other things including engagement, if we care to define what it actually means. (I suggest nothing like the 666 in the signpost at the entrance of the village). Empowerment, accountability, trust, and the rest are outputs, not inputs. It is what you get, not what you put in.

When it comes to culture, there is a lot of confusion in management thinking between inputs and outputs. For me, “what we put in” in behaviours. Behaviours are the input. I am not saying they are the only input. There is also attitudes, beliefs, values and mindsets for example. the problem is that I am no closer to knowing what to do with attitudes, beliefs, values and mindsets, all concepts in need of translation. Behaviours is something I can see, approve, like, dislike, promote, reinforce and above all describe in a very operational way. Behaviours create culture. Change behaviours, get culture change. This is the one-liner philosophy behind Viral Change

Culture, engagement, and all the ecosystem of concepts (satisfaction, productivity etc) are connected and entangled. It is important to define the conversation upfront to be able to make sense. There are also implications for how different corporate functions are involved in this ecosystem, corporate and internal communications being, together with HR, critical stakeholders. Mike Klein among others has been and is an expert and a voice of reason in this area of employee communications and engagement. The key learning for me is that it’s impossible to take a single angle or view, whatever that may be, and extrapolate to the whole broad issue of how employees experience the sense of belonging to a firm and their ‘alignment and engagement’. At the very least, any conversation needs to start with a dictionary not with a thermometer.

Download the full article here:

The corporate love affair with a thermometer – Employee Engagement surveys – Dr Leandro Herrero

© Leandro Herrero 2022

Dr Leandro Herrero is the CEO and Chief Organization Architect at The Chalfont Project, a UK based international firm of organization architects, specialised in large scale behavioural and cultural change, Viral Change™, Smart organizational design and Collective Leadership. He is the pioneer of Viral Change™, a people Mobilizing Platform, a methodology that delivers large scale behavioural and cultural change in organizations, which creates lasting capacity for changeability. He can be reached at: