We're a company manufacturing moulds and other tools that our customers then use in their machines. I was given a set of requirements for building a web application, which someone thinks that our customers would want to use.

Doing requirements engineering / BA work / stakeholder analysis, I wanted to stand back and first identify the business problem that the proposed application should solve. The thing is: Nobody can tell me. No customer survey was conducted, sales are unaware of problems the proposed web application could potentially solve, marketing hasn't identified a need either, the PM only sketches functional requirements instead of identifying the underlying issue (if there's any).

I have come to the conclusion that creating a web application for our customers is nothing more than an idea of one or more (probably high-level) stakeholder(s).

I think that the top managers would be available for a meeting where I could potentially inquire about the problem, but:
Wouldn't that potentially annoy those who initially proposed the development of such an application?

  • 3
    Edited the first part to clarify. The requirements all come from within our company, and that's the issue: Our own people talk about creating a web application whose intended audience is our customer, yet no one is aware that the customer actually wants / needs such a thing and it doesn't add value to our internal processes either.
    – Jan
    Apr 28, 2015 at 15:51

4 Answers 4


So if I understand right, your company thinks you should build an application for your customer. The customer hasn't requested it and you have no validation that the problem being solved even exists?

How you proceed is really going to depend on where you fit in the overall organization. I'm going to assume you're a program/ project manager (given what this subject of this StackExchange is).

I'm also assuming there is no portfolio level structure in your company. This would be the best way to deal with this. That's not going to solve this problem though.

If the requirements are coming from a product manager who is your peer or roughly equivalent to you: proceed with planning, so you can figure out the level of resources that would be needed. Then use the same guidance as below, for if the requirements come from someone senior.

If the requirements are coming from someone in a senior position (two or more levels above you): You need to practice the Pothole School of Project Management. In a nutshell, you need to find someone with the authority and interest to brief. For example, if an engineering VP came up with the requirements, try and bring this to the attention of the sales VP. The sales guy doesn't want to waste resources on something the customer doesn't want and also doesn't want to upset the customer by losing focus or delivering something they don't want.

The reason for this is as a project manager (program manager, scrum master, etc), you have no direct authority. You need to work with influence. If you try and tackle it directly, good odds you will be the one run over by the bus. The project management function is rarely seeing as having "expertise" so when an issue comes up, they are already at a disadvantage.

And as Tobias said, don't engage the customer. Even if you have direct relationships, you would be discussing internal company information and that's going to impact your credibility with the customer and your company.

Best of luck, would love to hear how it goes.

  • I had a hard time accepting an answer, since both seem to be correct. I managed to eilicit some requirements and trace them back to the percieved problems. To make sure I got the requirements right, I verified the problems and requirements with others in my organization. My next task is to point out that only a single person is aware of the underlying problems - or that this perception is wrong and there aren't any (at least regarding this case).
    – Jan
    May 6, 2015 at 11:19

Calling the top-management sounds pretty much like ignoring someone in the hierachy. If the PM is able to come up with functional requirements he has at least some business problems or market chances in mind - even if the PM is not able to formulate them, e.g.

  • The customer would like to use a fancy tool in oder to "feel" cool.
  • The company has to create a high-tech image (by using web technology).
  • Someone might try to sell fridges to the south pole (why not?)

Do the following in order to identify the business need:

  1. Take care that you identified all of the internal stakeholder.
  2. Identify the internal source of the idea
  3. Check "knowledge elicitation" techniques, e.g. create Ishikawa diagrams with several stakeholders
  4. Make your sponsor proud of you by fulfilling you task (identify requirements?) but highlight the identified risk of missing customer demands

The thing is: Nobody can tell me.

It's called innovation. Change your angle of attack. You seem driven to identify use-cases etc., however quite often, all this arrives in the flow of the project. This is also the territory of TDD, test-driven development.

What you should do, is this. You must keep all stakeholders informed and try to bring them to a round table from time to time if possible. Maintain a health schedule and documentation of your work, but don't over-work it. Understand how your sponsor sees your role and acting. She needs to understand risks and impacts from unclear and false requirements, etc. Good luck.

Edit: Highly recommended reading the materials by Karl Wiegers.


Creating a template or finding an example of well-written requirements to share with team members helps to establish minimum quality and consistency across the board. In addition, creating a glossary can provide the ability to reject poorly written stories based on a shared vocabulary. If all else fails, paper prototyping an implementation based on the requirements can demonstrate the gaps and motivate the team to help.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.