Tester Motivation – the new and the old (your chance to provide evidence – see the survey at end)
by Stuart Reid, TSG & Tafline Murnane, K. J. Ross & Associates
If someone told you that a container of hot water would freeze quicker than a similar container of cold water put side-by-side in a freezer, would you believe them? Surely, you would argue, the hotter water has to cool down to the temperature of the cooler water before it freezes, so it must take longer. But you would be wrong. In practice it works the other way round, which, for most of us, is completely counter-intuitive. If, like most testers, you want to see evidence to support this claim, then look up the ‘Mpemba Effect’ and you will see that this phenomenon has been documented since the time of Aristotle, circa 350BC.
So, how is this related to tester motivation? If someone told you that the effect on a tester of rewarding better-than-average performance with a bonus was to reduce their productivity, would you believe them? In the past the traditional ‘common sense’ approach of rewarding good performance and punishing unwanted behaviour was assumed to be the most effective motivator, but today’s experts tell us that bonuses often do more harm than good. In practice, promising (and giving) bonuses for creative work that requires the use of cognitive skills, such as test case design, actually de-motivates - another example of observable practice being completely counter-intuitive to most of us. Of course, discussions on bonuses and motivation only become relevant once an acceptable base level of income is first achieved. Nonetheless, even in the presence of a lower salary, evidence indicates that bonuses can still de-motivate employees.
As a manager it is difficult to over-estimate the importance of inspiring testers to achieve the goals of the organization; testers who are motivated are more likely to strive for higher quality outcomes, achieve objectives and work in a more efficient manner. Motivation and workplace satisfaction are not, however, simply a managerial concern. If you are a test practitioner, you spend a large proportion of your waking hours at work. There can be few more important things in your life than to ensure that you feel motivated by your work.
Basing our style of motivation on management theories taught to us at university or on the approach used by our present or previous managers may mean we are missing out on the optimal approach, as this means we may be ignoring more current theories. Working as testers, at the heart of the IT discipline, we should be accustomed to technological advances changing how we think and work on a regular basis. The field of motivation has seen a similar rate of evolution over the last 30 years, which means that we need to be aware of the current state of ‘best practice’ and identify how this can be applied to software testing.
Currently, the three major factors thought to determine motivation are autonomy, purpose and mastery.
Autonomy takes a variety of forms, but typically includes the tester’s ability to decide which activity they perform next, how they perform it, and, perhaps most importantly for testers, the time and quality associated with the activity. Autonomy can work at many levels. For the individual who performs the testing, there are two possibilities – either they decide what type of testing is appropriate, or this decision is imposed on them by their test manager or test lead. At a project level autonomy is often equated with the principle of empowerment within agile teams. In this situation we would expect to see testers able to decide (along with the rest of the agile team) on aspects such as the appropriate levels for test completion criteria and which tools to employ. In reality, we often see that it is difficult for managers to relinquish their day-to-day control to the agile team and so often the expected benefits of team self-management and optimisation are not always achieved.
Some organizations have started to implement a wider form of autonomy by providing ‘free time’, allowing their employees to spend a proportion of their time on projects that they really want to do. Google (20%), 3M (15%) and Atlassian (20%) are high-profile examples of organizations using this strategy. At first glance this appears to be rather indulgent and perhaps only possible for organizations that are particularly ‘resource-rich’. A closer look, however, contradicts the idea that this is an extravagance when the benefits are considered. First, such a strategy contributes greatly in persuading staff that their ideas are valued thus increasing the retention rate and encouraging a flow of good new candidates into the organization. Second, this strategy is really successful at generating income. Most of us use Post-It Notes, a result of the 3M 15% rule, and Google estimate that approximately 50% of their new products come out of their 20% innovation time (one example, AdSense, currently generates approximately 30% of their revenue).
Some testers, when offered the opportunity to choose what to do with a portion of their work week, decide to work on not-for-profit projects for the benefit of the wider community rather than devote this time to increasing their employer’s profits. This introduces the second motivation factor of ‘purpose’. Examples of activities undertaken by testers in this area are open source development and testing, organizing peer groups for testers, and contributing to tester certification schemes and testing standards.
Of course, having a job with ‘purpose’ does not mean that your employer has to give you free time in which to do it separately from your main role. For many testers their jobs can be inherently purposeful. They may work for a charitable organization, or they may feel completely aligned with their employer’s mission (e.g. developing environmentally-friendly products). In contrast, when testers find themselves testing a feature where they are unaware of the purpose or (even worse) with which they disagree, then their motivation is very likely to be undermined.
Having a job with autonomy and purpose is great, but these factors need to be complemented by the recognition that nearly everyone also wants their job to provide a challenge. Being given the freedom to choose when to run tests on an application that you feel is really worthwhile is great, but the motivation will soon fade if the testing involves no intellectual challenge. The ‘mastery’ factor introduces the concept that we want to be challenged and to get better at what we do. We want a target of mastering a skill, albeit with the understanding that in reality we will never actually achieve full mastery (as we get better, we will identify further gaps in our knowledge and so identify more skills to master).
Another way of thinking of mastery is the concept of feeling you are ‘in the flow’ (or ‘in the zone’). In this state you are constantly pushing at the boundaries of your skill, but never to the extent that you feel overwhelmed by the challenge. Thus your mastery of the skill grows, and, ideally, this growth should be an end in itself. Typically if you are in the zone (and undisturbed) then time flies by without you noticing. Mastery comprises a number of aspects, and would ideally require individual testers to decide what skills and goals they wish to attain, and also involve them in designing and optimising tasks that will better keep them ‘in the flow’.
So far we have considered autonomy, purpose and mastery as the three major factors currently thought to determine job motivation. Back in the 1970s, the Job Characteristics Model provided a well-respected and widely-used approach to the design/re-design of jobs to achieve both motivation and satisfaction for the employee and higher retention and productivity rates for the employer. The model is based on the idea that five job attributes can be used to provide a measure of how satisfying a job is. Scores are recorded for each of the five attributes and combined to produce what is known as the motivating potential score (MPS) for the job.
The five attributes, scored in the range 1 (low) to 7 (high) are:
Skill Variety (V) - the range of different skills needed;
Task Identity (I) - the degree of completing a whole job;
Task Significance (S) - the importance of the job, as perceived by the tester;
Autonomy (A) - the level of control of their own time;
Feedback (F) - the degree of supervisory and results-based feedback.
MPS can then be calculated using the following formula:
As can be seen, autonomy and feedback are the most important factors accorded by the MPS formula, but mastery and purpose are not included in this earlier (and tried and tested) approach to motivation. Thus it can be seen that in the last 30 years there has been a revolutionary shift (around the autonomy axis) in which are the most important factors for job motivation.
A number of previous surveys have been performed that consider the application of MPS to various IT jobs, including testing (tester’s jobs typically have the lowest motivation scores in IT!), but as far as we are aware no empirical evidence has been gathered to date to provide insight into the validity of the new autonomy, purpose and mastery model of motivation as it applies to testers. To this end, a survey has been designed to collect data to allow the two models to be compared with each other and with practitioners’ real-world experiences. The aim is to analyse the results to provide insight into how we can both improve our own jobs as testers and create motivating jobs for the testers we manage.As a test practitioner, we invite you to take part in the Tester Motivation Survey. It should take no longer than 15 minutes. All participants will be provided with a copy of the resultant white paper on Tester Motivation and they will also be offered feedback on their personal results, if they so wish. All input will be treated as confidential and no individual survey data will be released to anyone other than the original participant.