7 mins to read

Listen carefully: Please kill the engagement survey – Part II

A 20th-century management style keeps coming back.

Mon, 09 Jan 2017

See Part 1 here

Interactions between people can’t be deconstructed and presented as pie graphs. 

Human interactions are complex. More specifically, they are complex adaptive systems. The study of complex adaptive systems is an ecological metaphor as opposed to a mechanistic metaphor. 

Interactions between people change the system and the system changes the people, continuously. This gives rise to emergent behaviour and attitudes. Emergent behaviour is by definition non-linear. Small changes can have large impacts on engagement, and conversely, large changes may only have small impacts. 

Parts of the organisation may change significantly while others remain fixed. The system evolves but seldom repeats, what worked yesterday might not work tomorrow. Complex systems cannot be deconstructed into their component parts. The system is an emergent quality of the interaction between the parts not a property of the parts themselves.

Engagement is an amorphous concept. We all have a view of engagement but experience it differently. Our level of engagement is constantly changing and is influenced by a variety of factors both inside and outside our role or workplace.

Placing a numerical value on an amorphous concept (the engagement index) has the same validity as placing a numerical value on how much you love your partner, or finding a coefficient for how friendly your neighbours are.

Our personal engagement is an emergent quality of the system. We cannot take a scientific management approach, dissect it into its component parts, fix it and reassemble. 

Trying to analyse a complex system in terms of statistical modelling has limited value. In a continuously changing dynamic system, knowing the mean value of a set of arbitrary factors at a specific point in time returns almost no useful information. It is literally a waste of time.  

Using this information to benchmark against other organisations is pseudoscience. No two organisations are the same. Context is paramount. Using the data to identify annual change in engagement is equally silly. In a system that is constantly changing, comparing the results of two arbitrary snapshots is completely meaningless. 

Engagement surveys are often used as a proxy for intimacy. We don’t know how people feel so we survey them. If you want to know how people feel, ask them. They will tell you.

The trouble with survey
Most surveys are inherently flawed, even when they are used in an authentic manner. 

Many telegraph the “right” answer. Employees who feel happy will give positive answers and employees who feel upset will punish the establishment by giving negative answers. A question such as: “does the person I report to treat people with respect?” is an obvious candidate for emotional bias.

The bias is compounded when people start to understand that management and in some cases the board of directors will hold people to account for the result. So they actively game the system.

People inherently understand that human relationships are complex and hold little regard for the engagement survey. We all know the mood of an individual or group of individuals can easily be understood but cannot be expressed in numerical terms. Maybe we should stop pretending?

Take a look at two recent cases where predetermined assumptions were shown in retrospect to be poor indicators of the mood of the people.

In the UK Brexit vote and in the US election, polling experts used an assumptions-based model to convert a sample poll into an election prediction. In both cases they were woefully inaccurate. 

The UK and US were facing unprecedented times. Past voter behaviour would be a poor indicator of current voter sentiment. The assumptions made were based on historical information and the result was both unpredictable and surprising. 

An organisation is constantly changing. Complex systems will always return unpredictable results. Distilling interactions into a set of factors and measuring them at a fixed point in time against predetermined success factors starts to sound a little silly, yet with engagement surveys we are doing just that.

The danger comes when we take the mean result and use it as a KPI. It turns out that most people resent being held to account for the outcome of a process they reluctantly participated in, is logically flawed and is actively gamed. 

Surveying engagement actually decreases engagement. So if we can’t numerically quantify engagement and we can’t measure engagement. What can we do?

Treat people like people – actively listen to your staff.
As human beings, we are tuned into each other’s emotional state, this is a feature of our evolutionary success.

Because of this, we intuitively know who in the organisation is engaged or satisfied with their work and who is not. We either tell stories about our daily experiences or we signal our emotion through our body language or behaviours (such as mediocre results, or not participating in teamwork).

Humans don’t express amorphous concepts like engagement by evaluating them. If someone asks you “I was thinking of applying for a job at your company, what is it like is to work here?” you wouldn’t provide an evaluative answer “Our latest engagement index was 68.2% up from 63.8% last year but still well below the national average of 77.5%”. 

You would relay a descriptive story or anecdote that expressed the complex concept of engagement. We are adept at communicating and remembering information in story form. Stories naturally involve cause and effect, example and explanation, some reality and some exaggeration. 

If you can harness people’s natural desire to express through descriptive anecdotes, then you can use the natural occurring narrative within the organisation to provide opportunities for improving engagement. We can do this without having to quantify it, and without having to get experts to tell us what it means.

Importantly, we can (and should) do this at a granular level, the mean for the whole system is not important. Rather than looking for whole-of-organisation solutions, empower people to increase engagement within their own span of control.

Develop two or three open statements that direct attention toward the subject. Employee engagement might be covered in two questions.

For example:
a.    “Tell me a story about a time when you felt really demotivated at work, what caused that and what was the outcome?”
b.    “Tell me another story about a time when you felt really motivated, what caused that and what was the outcome”

It may be important to have some objectivity in this narrative gathering process. People will tell stories differently depending on their relationship to the recipient. Also, get people to tell stories in ways that prevent individuals being identified. There is software available to record stories without an interviewer present.

These narratives can be transcribed directly without losing context. A benefit of this process is that people can see stories that make them both proud and embarrassed. Humans remember stories. Data doesn’t carry this emotional power. 

In the workshop, ask the team to seek patterns in the narrative. Patterns can be opportunities for innovation, overall themes for improvement or areas of strength. Ask the team to identify opportunities for immediate operational improvement.

Finally, ask team members to summarise solutions to be actionable, practical and pragmatic. Implement those solutions immediately. This shows staff you are focused on their engagement and emotional welfare. It will also show how honest conversations can be translated into action quickly and effectively.

Leading Change
Most organisations are designed as hierarchies. In a hierarchy, management can control the system and impose its will by setting rules and enforcing compliance. 

In a network, the manager is a node. The network is governed by an ever-changing mixture of constraints and boundaries. You cannot direct a network, only influence it. Rules-based systems are usually too rigid to function well. People will break the rules to get the job done. 

As an example, how long would it take for your organisation to cease functioning if everyone did exactly what was on their job description and nothing else? Work to rule is a threat not a promise. 

In a network, success is in the interactions not the plan or the rules. This means the role of a manager in increasing engagement is:

Stop using survey as a proxy for intimacy – talk to people
Listen to the naturally occurring narrative – continuously, not periodically
Act immediately and at a local level (your span of control) 

Instead, ask:
1.    What can I do right now to increase the positive stories I am hearing about engagement?
2.    What can I do to right now disrupt or mitigate the negative stories I am hearing about engagement? 

It is what you do that matters, not what you say. The result of a workshop on engagement should be a clear set of actions not a plan – or a plan to make a plan. It should never include a counter-narrative about why the engagement narrative was wrong or misguided. That is similar to telling an upset friend they’re not really upset, just misguided.

Never put KPI’s or targets around engagement, people are not cogs, the system will never stabilise to a point where you have succeeded or met your goal. People will simply game the system and be annoyed they had to game the system.

Increasing engagement is a continuous process of active listening, continually amplifying the positive and disrupting the negative.

What you do is far more important than what you say. Quantifying engagement is not necessary or desirable. We all know positive engagement when we see it. 

Kill the survey, just talk to people.

Steve McCrone is managing director of Cornwall Strategic

© All content copyright NBR. Do not reproduce in any form without permission, even if you have a paid subscription.
Listen carefully: Please kill the engagement survey – Part II