Employee engagement – is performance poor without it?

Written by: Sarah Page-Jones
Published on: 5 Feb 2021

Vet with cat

Work enjoyment impacts job satisfaction.

Most of us will have encountered employee engagement (EE) surveys because businesses believe more engaged employees perform better.

A common definition of an engaged employee is a person committed to the organisation and motivated to work for it. But huge definition variation exists and, sometimes, those selling engagement surveys and their customer organisations never clarify whether they are talking about engagement in the same terms.

Most businesses now measure EE, and spend considerable money, time and effort on surveys and projects aimed at improving engagement in the hope of increasing revenue.

The Harvard Business Review (HBR; 2013) report on the impact of EE on performance opens with: “Employee engagement has become a top business priority for senior executives… They recognise that a highly engaged workforce can increase innovation, productivity and bottom-line performance.”

Links to performance

With an ongoing focus on EE and high-profile organisations speaking passionately about the benefits, it is natural to believe we should all concentrate on it. The assumption is an engaged workforce performs better for the organisation (note HBR also assumes better innovation), but is this true? Or simply what we are told by those selling EE surveys?

EE surveys tend to measure job satisfaction (JS) and organisation commitment (OC) – both of which have a weak positive link to performance.

A meta-analysis by Riketta (2008) looked at longitudinal studies to see if increases in JS and OC lead to an increase in performance. He found that satisfied and committed employees only perform a little better than their grumpier counterparts (where OS was a slightly better predictor of performance than JS). The longer the gap between measuring EE (JS and OC) and performance, the weaker the link becomes. For example, a 1 unit increase in EE leads to a 0.12 unit increase in (both task-based and extra-role) performance for a 1 to 6-month gap and a 0.02 unit increase for a 7 to 12-month gap. In short, changes in JS and OC result in incredibly small changes in performance over time (so the answer to our title question is an emphatic “no”).

Also, little support exists for the reverse effect (that is, performance driving JS and OC), but the small amount of evidence available suggests good performance is more likely to lead to better JS. In short, you are more likely to be satisfied with your job if you earn a bonus each month.

Most companies offering EE surveys measure all sorts of parameters under the umbrella of EE, and this differs between companies, so it is hard to compare products or results. EE companies tend not to investigate whether a link exists between engagement and performance in an organisation. In addition, EE is rarely defined; is it JS or OC, or something else, such as work conditions, or a combination of these and other factors? Also, the meaning of performance is seldom made explicit, but is usually assumed to be financial.

EE companies tend to measure what they think relates to engagement and proceed to tell organisations what must be worked on to make engagement “better” (with the assumption that performance will then be “better”, too). The “proof” of impressive outcomes is often based on anecdotal case studies and the “expert” experience of the consultant or salesman. We believe it because it sounds intuitive.

Hidden variables

If EE survey companies do measure engagement and performance, they tend to do so simultaneously rather than longitudinally and then assume one has led to the other. Even if EE and performance appear to correlate when measured at the same point in time, this does not mean better EE leads to better performance; it may be the other way around, or a hidden third variable may be involved.

For example, a correlation exists between drowning and good ice cream sales. This does not mean more drowning leads to better ice cream shop performance (or vice versa) – the hidden third variable is the ambient temperature. When it is hot, more people get into trouble swimming in the sea and more people buy ice cream.

The two variables happen to be high (or low) together – no causal link exists.


What about the surveys themselves? If you want to proceed with a questionnaire for reasons other than improving performance, you might want to consider a couple of other things before investing.

Firstly, would the top management team be willing to make any necessary changes? If not, it makes little sense to proceed other than for superficial employee relations and this may be detrimental to the business. For example, if you signal to employees you care about their JS and then ignore the survey results, disengagement may increase.

Secondly, we must think about the questions being asked. What do they cover (that is, what does the EE company include in its definition of engagement) and are they relevant to JS or OC in your business?

For example, the author saw in an EE survey a question asking front line employees if they understood the wider organisation’s funding structure. Most of them did not and so this was flagged by the EE company as an area requiring improvement.

But does it matter if front-line staff do not understand the full range of business income streams? Will it make any difference to them or the business if they do? Probably not. Consequently, we must be careful not to invest in improving something that might be a “nice to have” just because we are told we must. Furthermore, it is rare for the results of an EE survey to be a complete surprise to managers. Most know something about their team member’s JS and OC without the need for a survey (even if they do not want to admit it).

So, think about the questions that will help you understand areas you are not sure of or explore known issues better.

We must also consider bias in response rates. If your response rate is low, you cannot extrapolate results (no matter how tempting) and you need to think about who is responding. For example, if most of your responders are managers rather than front-line workers, you might pat yourself on the back because your team is engaged, but the people who earn the revenue might be so disengaged they cannot be bothered to fill in the questionnaire.


In conclusion, if you want to increase performance in your business, do not waste time, money and effort on EE in the hope of a noticeable increase in revenue. Define the type of performance you want to improve and focus on that directly.

While doing this, remember performance is a complex phenomenon (it depends on job type and context) and is hard to measure accurately (see article two in this series; VT50.51), so performance improvement plans take some thought.

Measuring EE may be helpful for your organisation – for example, in managing employee turnover (this will be expanded on in the final article of this series) or absence rates, and to assist with improving employee well-being. But before you start, make sure you know what you mean by EE (JS and OC) and the questions you ask are targeted appropriately. You can then evaluate the results effectively and make informed business decisions.

Do not be afraid to ask for the evidence behind claims EE is the key to organisational success and do not be swayed by persuasive anecdotal case studies, testimonials or what other prominent businesses are doing. You will be better off if you go against the crowd because EE surveys only make a significant difference to the financial performance of the companies that sell them.

  • This article was first published in Vet Times (Volume 51, Issue 1, Page 12-13).


Harvard Business Review (2013). The impact of employee engagement on performance, Harvard Business Review Analytic Services Report:1-16.

Riketta M (2008). The causal relation between job attitudes and performance, Journal of Applied Psychology 93(2): 472-481.