Asking requires listening and writing down what you hear!

Asking requires listening and writing down what you hear!

Asking stakeholders to describe or define requirements is the most common way to develop requirements for projects. Specific techniques include talking to stakeholders in the hall informally, interviews and questionnaires and very formal Joint Application Design (JAD). These techniques are popular because asking and talking to people is easy and opens a dialog. However, while stakeholders may know their business need, they may not know the details of what they really want and need. Moderation and planning are critical for making all of the techniques in the category as effective as possible for creating an initial backlog. Examples of how moderation and planning could be implemented in two classic “asking” techniques are shown below:

Joint Application Design (JAD) is a very formal technique that is an off-shoot of Joint Application Development that evolved in the 1970s. JAD is highly structured approach to developing the requirements and design for an application or project. The process is based on the interaction between key roles (sponsor, subject matter experts including business and IT participants, facilitator, scribe and potentially observers). The process requires all roles. It should be noted that the JAD process was one of the earliest techniques used to embed business and IT personnel for any substantial period of time. The process (documented many places including Wikipedia) has a number of key steps that provide a structured approach for interaction and generating information. Setting the goal (one of the key steps) of the JAD acts as an anchor for the process and provides a tool for the facilitator to re-focus the process if it wanders off course. In order for a JAD to work, up-front planning is mandatory. The participants need to be carefully identified, the goals of the JAD identified and a detailed agenda with supporting documentation needs to be developed. Preparing for the JAD can take as long as the session itself. JADs typically run three to eight days and participants typically were sequestered from the typical working environment during the session.   The combination of skilled facilitator and structure help IT and business participants interact in a creative and productive fashion. Overall JAD is a very powerful technique, however the structure and overhead tend to make it more difficult to apply.

In its classic form, JAD is viewed as less than Agile. Historically it was used to develop the much abused, big up-front design (BUFD). Agile principles call out the concept of emergent design, while eschewing the BUFD. The practice of Agile  is generally more a reflection of finding the balance between what needs to be known and what needs to be discovered. I have used the formal structure of the JAD process as a tool to initiate Agile projects very successfully by refocusing the goal to build an initial product backlog. The combination of structure and facilitation is more valuable when a team is addressing a new business area or in matrix organizations where teams are assembled for each new project.

Interviews are another of the classic “asking” techniques. Interview techniques can range from formally scripted question and answer sessions to loosely guided discussions. Formal interview techniques begin by developing a set of questions to be asked during the interview. In formal interview situations, the responses to the questions in the scrip and any follow-on questions captured as close to verbatim as possible. A legal disposition is an example of a formal interview. They require the interviewers to prepare for the interview not only by developing the set of questions to be asked, but also to gather information about the general outline of the answer they are going to receive. A good interviewer is rarely surprised by the answer they receive. Informal interviews are typically less structured, however they still require preparation. In less formal scenarios I generally recommend developing a loose set of framing questions (framing questions capture the direction of interview without being specific) so that the interviewer develops a goal for the interview and then plans the approach to attain that goal. The framing process is important in case the interviewee throws a curve so that interviewer can gradually guide the interview back to the correct track. Take notes (do not trust your memory) in all interviews. While informal interview seem more like common conversation, interviewers that are good at the informal technique tend to good counter-punchers (able to deliver well formed follow questions that keep the interviewee talking) however even in an informal interview, the interviewee must always their ultimate goal in mind. In both formal and informal situations, if the interviewer is emotionally involved in what the answer should be, consider using a facilitator or external interviewer.

Asking stakeholders for requirements is a tried and true method to generate an initial backlog. Asking should not equate to ad hoc or mere order taking. Asking requires preparation to be effective whether using formal techniques based on JAD or informal interviews. As an interviewer you need to map out where you want the session to go and then act as the guide.

The backlog represents blooming ideas

The backlog represents blooming ideas

Many discussions of product backlogs begin with the assumption that a backlog exists, but where does it come from? Whether you are building a new application or starting a new enhancement project, you need an initial backlog and many times that means that it needs to be developed. In most cases the initial backlog is not going to be delivered by the tooth fairy or by the product owner. Generating the initial backlog requires an investment of time and effort by the project team. Developing the user stories for the initial backlog can be accomplished in many ways. A few common approaches for building an initial product backlog include:

  1. Interview/Discussion Based – The first group of methods boil down to talking to the people that use the current system or know the business needs for a future system. Interview and discussion techniques gather people together and draw information out. These types of techniques have the benefit of being easily scalable. Interviews and discussions can be done one-on-one, small groups or in large groups.  Groups generally leverage techniques for facilitated sessions such as affinity diagraming or Joint Application Design (JAD).
    All interviews and discussion techniques require some degree of preparation. In informal interviews the interviewee needs to understand the business well enough to ask questions and should have both a goal and a strategy for the interview. Develop a question set for more formal sessions. A defined question set is a necessity for large groups or it is too easy to not get the information you need.  I have used variants of the killer innovations questions found Phil McKinney’s book Beyond The Obvious as an interview framework. The facilitated process detailed in Ellen Gottesdiener and Mary Gormans’s book Discover to Deliver (my review and interview with Ellen and Mary) along with affinity diagraming are two techniques I have found very useful for guiding facilitated sessions.
  2. Observation/Documentation Based – Observation and documentation techniques are very useful in scenarios where there is a current manual process that is going to be automated or where interviews are difficult to schedule. Observation allows the observer to watch how work is done with the benefit of not having to rely on what can be described in an interview.  Interviews are subject to interpretation and cognitive biases. The downside to observation is that it is difficult to scale and is most effective when you use trained observers. Documentation techniques done by reading and reviewing system and user documentation, code and other system artifacts (like user interfaces and reports). [FRAGMENT] The review generates the requirements for the new project.

The goal of interviews, reading documentation or observing people working is to generate an initial list of work items. In most cases we are targeting getting just enough of the stories to get the project moving. The work items that are captured need to formatted and groomed just like any other backlog item. Backlog items should be formatted using the standard user story pattern (persona – goal – benefit), should be estimated and should have acceptance criteria.

Where do you get your initial product backlog?  The process begins by collecting needs and requirements.  These needs and requirements will be both business and technical nature.  Once you have this data you can apply standard user story grooming techniques.  For example, we can use story mapping to help drive the needs collection by mapping the information in order to expose gaps.  All of these techniques require one basic input – business needs and requirements.

Happy New Year!

Happy New Year!

It is that time of year, part 2! Today we celebrate the ten top interviews from 20013.  I learned something from every interview I did this year.  They all contain nuggets of pure gold that changed how I worked, and in some cases changed how I lived my life. While all of the interviews were great, today I want to share the top ten downloaded interviews from the 2013 ‘ish.  The top ten most downloaded interviews were:

SPaMCAST 254 – Matt Heusser, Agile Testing, Test Professionalism
SPaMCAST 220 – Stephen Parry, Adaptive Thinking, Command and Control
SPaMCAST 264 – Alexei Zheglov, Lean and More
SPaMCAST 250 – Ben Linders, Success With Agile
SPaMCAST 246 – Tobias Mayer, The People’s Scrum
SPaMCAST 224 – Mike Burrows, Kanban Values
SPaMCAST 236 – Taylor, Rosenhead, James, Strategies for Project Sponsorship
SPaMCAST 248 – Pat O’Toole, Project Improvement, Process Improvement
SPaMCAST 258 – Steve Tendon, Hyper-productivity
SPaMCAST 256 – Kenny Rubin, Scrum, Economic Frameworks for Agile

Click on the link to listen to any that may have missed or just to treat yourself to another listen.

Any trends?  The only trend I see is that SPaMCAST listeners love an eclectic mix of new ideas regardless of framework or methodology.  The top ten downloaded interviews in 2013 were a cross section of tips for immediate consumption and also about building base knowledge about how work is and can be done.

I hope your holiday season met your expectations and that you have set your goals for 2014 astronomically high. The one thing I know about my listeners is that they are all overachievers.  Let me know how the Software Process and Measurement Cast can help you meet those goals!

Notes:  I complied download statistics for podcast released between 1 December 2012 and 30 November 2013.  SPaMCAST specials, for example the downloadable copy of the SNAP Counting Manual, were not included.