Experience Sharing from colleagues: Share a case study of demand research and talk about his feelings and Inspiration

In this project, the demand investigator spent one and a half months opening eight demand exchange meetings and sending more than a dozen demand discussion emails to the subject to investigate the demand, three editions of the Requirement Specification (all over 60 pages) were generated during this period ).

After some system functions are developed, they will be displayed to the owners. After reading this, the subject exclaimed that it was not what they wanted and accused us of not developing as needed. After we took out the requirement specification, the subject was even more excited that we "tampered with the demand ".

After that, the Project Manager explained to the owners from the business point of view why the system was implemented and how the system would be problematic according to the ideas of the owners, the owner accepted our statement and failed to pursue it.

Some people will say that the main problem in this case is the problem of the owner. The demand survey sent three versions in a month and asked them to confirm the problem. They did not look at the problem behind the demand, but they also insisted on it. However, do you think we have any shortcomings? My opinion is as follows:

1. During the survey, the requirement specification is directly provided to the owners, which is the practice of "Overlord hard bow. Some people may say that a survey has talked a lot about the content. It is better to write the requirement specification than to write the meeting minutes. It improves the efficiency in one breath. Why not? Dear friends, do you have any questions from the perspective of owners? How long can we concentrate on a 60-page document? How many times can I see it? According to my practice, we can't concentrate on around 20 pages. How can we ask the owners if they have problems in their own work? However, the owner is not responsible, and he cannot trust you.

2. Taking the requirement specification as the only output of the requirement survey is to put all the eggs in one basket. After the event, the Project Manager carefully searched for the requirements survey meeting minutes. The requirement confirmation email could be the confirmation content without the requirements details. The results of the Demand Survey do not allow the owners to confirm each point. It is the opportunity to "bury mines" on the Key Path of the project to the owners challenge. This time, the mine was not exploding, will it be so lucky next time?

3. After talking about the problem, let's talk about the improvement methods.

Reach an agreement with the subject regarding the process control of the Demand Survey. Not only do we need to understand, but also let the owners understand that the requirement specification is the final result of the requirement survey and is not the only result. The content of the requirement specification must be approved by the owners (preferably signed for approval ); then we need to reach a consensus on the validation cycle. For example, the subject must make a check for the survey requirements one day ago, our responsibility is to issue the demand survey content for the day in the form of meeting minutes within the day. A detailed update list and requirement SRs should be provided to the owners after the completion of the phased survey to facilitate reading.

To gain the trust of the owners, we should not only understand the business, but also look at the way we do things is not professional enough. Is it professional? The comparison of professionalism is reflected in our process control details. As long as we want to know before the owners, process control is more subtle than competitors, in the eyes of the owners, we are more professional than our competitors.

If you have said so much about the demand research, do you have any methods to control the research process?