If stakeholders are not familiar with requirement workshops, have no time to spend on them – or don’t want to spend time on them, then you can decide on development techniques for these stakeholder groups that demand readiness to communicate and commitment to a lesser extent. Observation techniques like field observation, apprenticing and contextual inquiry belong to these techniques, as do surveying techniques like interviews, questionnaires and self-descriptions, as well as supported techniques like persona scenarios and use case scenarios. Today, we will take a closer look at field observation, apprenticing and contextual inquiry.
In field observation, you watch over the shoulder of stakeholders at their workplace and so gain some knowledge of how the planned system should work. There are two variants of this technique:
- Ongoing observation,
- Work-sampling study.
Leading ongoing observation means to experience activities, the uses of work documents and materials, to experience the solid environment and working conditions, and to record essential impressions in text. The work-sampling study is a sampling by which statically secure dates and measurements are derived from a variety of momentary observations.
Through field observation you can gain a variety of diverse insights, about, for example:
- The practicality of workplaces,
- Unusual burdens in the workplace,
- Occupancy rate of the workplace,
- Time required for processes,
- Tight situations,
- Communication paths (and shortcuts),
Tips for field observation:
- Determine in advance what you want to observe, so people, workplaces and things – looking at the system being developed – that could be of interest.
- If you want to carry out a work-sampling study and the result should have static reliability, you have to calculate a required number of observations.
- Create an observation schedule and confirm with those concerned.
- Notify the concerned persons. Because there are often aversions to the observation technique, those concerned should be precisely informed about the purpose and the crew of the observation and be made aware of the appointments well in advance. Attention: as tempting as it may be to use your smart phone for support (voice recordings, photos, videos), nothing is allowed without the consent of those you are observing, the management, and it could also fall under the duty of co-determination.
Our recommendation: Avoid ongoing observation. It is only a recommended technique in rare circumstances, because of its large time demand as well as the psychological burden on the stakeholder in question, and because there is a risk of false data that should not be underestimated: people change their behavior, consciously or unconsciously, when they know they are being watched. Work surveying avoids this danger, requires less effort, delivers good results in the framework of the probability theory and is also linked with fewer disturbances to the stakeholder with whom you are working.
How does apprenticing work?
With this technique, the requirements engineer learns the activity of the stakeholder from the stakeholder and then carries it out themselves. Behind this is the idea that doing someone oneself leads to knowledge more effectively than talking about the stakeholder’s task. The requirements engineer then recognizes details whilst working that would not have come up had they spoken about it. A new outsider to the field looks at the work from a different perspective than a stakeholder, who has perhaps been familiar with it for years. As a requirements engineer, lots of details will be perceived that the stakeholder thinks is obvious and therefore did not bother mentioning. If you are advised by a zealous teacher, then the large danger of this technique is to slip into too many details and exceptions.
Tips for the apprenticing approach:
- Determine which stakeholder you would like to be taught by. Not everyone is a good teacher. You will need the support of management to find a recommended person.
- Have a prior conversation with the person. Explain viewpoints and the process.
- Work out a study plan with the teacher: what should you learn in what time frame?
- Start your study program. It should take place in four phases:
- Learning by watching: The teacher demonstrates the work processes in real cases.
- Supported individual tasks: You try it out yourself, but the teacher is at your side to intervene if required.
- Working individually with decreasing support (fading): The more competent you become, the less the teacher intervenes.
- Coaching: The teacher just watches and corrects you in emergencies.
This technique goes back to Karen Holtzblatt and Hugh R. Beyer, two experts from the areas of the user-centered design¹. We have these two authors to thank for another surveillance technique:
What is a contextual inquiry?
It is something between an interview and an observation. Many authors speak of a structured survey because the stakeholders are first interviewed based on a list of standard questions. This doesn’t happen in a “neutral place” like a conference room, but directly in the working environment of the stakeholders. Then the stakeholders will be observed in their activities. During the observation, if necessary, they will be asked “what are you doing now, exactly?“, “what triggered that?“, and “where did you get this information from?”, etc. The goal is to recognize the need for change and potential for improvement and to be able to derive requirements from this.
The contextual inquiry technique is based on four principles:
- Focus. Make the purpose of the survey clear: what do you want to know, experience, achieve? Plan your questions and processes orientated toward this purpose. This is the best means of not losing yourself in detail. Because precisely that is the biggest danger with this technique.
- Context. Go to those surveyed at the workplace, experience the environment and watch the work processes. Establish how the cooperation with others goes, the problems or disturbances that arise at the workplace, and how they are dealt with.
- Partnership. Speak with the stakeholders about their work in a trusting way. Encourage them to name unspoken parts of the work. Let the work steps be shown and, better yet, try them out with the assistance of the stakeholders. That way you can, through your own experience, answer the question: Why like this and not another way?
- Interpretation. Don’t pull back at the end to think about your part. Instead, develop a common understanding of the aspects of the work that have been commonly recognized as difficult, complicated or in need of improvement.
Tips for approaching a contextual inquiry:
- Determine which stakeholder you want to accompany in their work.
- Carry out a prior conversation with everyone who you want to accompany in the workplace. No one likes being watched. So introduce yourself and explain your requests. Explain the process, the goal and the focus of the contextual inquiry. Determine, together with the stakeholders, which special work processes the stakeholders will carry out during the contextual inquiry.
- If you are allowed to use recording media, ensure the stakeholders absolute confidentiality.
- During the observation and interview phases, make notes about everything that happens to the stakeholders at the workplace – even the things that don’t necessarily belong to the work process.
- Summarize what you have seen and recognized, and discuss (in the sense of the principle of interpretation) what you have experienced with the stakeholder.
Because the stakeholder is asked in their trusted work environment, the contact threshold is low. As a requirement engineer you don’t only experience the technical and environmental, but also the social environment at the workplace. The impressions gained are more real, the collected information is more precise and easier to evaluate than the ones that are shared in the “lab situation” of an interview. With this technique you can gain detailed information. They assume a not-inconsiderable measure of empathy from your side.
When determining requirements, surveillance techniques offer a range of advantages. One of the large advantages is the recording of the actual behavior of the stakeholders. It is without euphemism and not linguistically upgraded. Through these social interactions between stakeholders, you can often gain additional knowledge. And the required cooperation of the trial is relatively small in comparison to other forms of deriving requirements. However, there is a high effort for the preparation, execution and evaluation of observation. The largest disadvantage is the fallibility of the watcher. Impressions are quickly evaluated differently, preconceptions confirmed or features systematically judged as too good or too bad. And maybe the stakeholder under observation changes their behavior and therefore delivers false results. My personal estimation, then, is: observation is a good instrument especially if the execution of requirement workshops is hard to realize. The interpretation of the results, however, can be tricky.
 Further information about apprenticing can be found in Beyer, H. R., & Holtzblatt, K. (May Vol. 38, No.5, 1995). Apprenticing with the Customer. Communications of the ACM.