What are effective strategies for dealing with project bottlenecks?

Question

I work as a software developer at a small non-IT company. One of the challenges we face is receiving timely feedback to questions and the testing of the programming assignments we handle. The root cause of this is that there is usually one project requestor per department who will review/test our software and this requestor will frequently have multiple projects they are involved in at any given time.

I am not sure if we have an employee resource allocation problem, or if a particular project management method might help the situation.

From the staffing side, we seem to have multiple programmers working on multiple projects for one requestor. At times, the requestor ends up very overwhelmed and feedback on a given project can take weeks to months to be received. Not a good situation for anyone. However, it seems there are plenty of times when it is possible for someone to handle the review/testing of multiple projects as not everything is ready for review simultaneously and things can progress forward.

On the project management side, we do not follow any true methodology, unless chaotic-waterfall is an approach. Would an agile/scrum methodology help this situation? I guess my confusion on the scrum approach is the allocation of someone to review/test the software produced and answer questions for the programmer.

I feel like my company could make some nice improvements to how we do things, but I am struggling with a direction for us to research/consider. Since we are a small company, and not software-focused, it is challenge to find the best improvements to make. If anyone can provide some direction, it would be greatly appreciated.

Answer 1

I actually think that this may be a classic case of a missing "Product Owner" or "Customer". No process will help you with this. I think you already identified the issue that the person reviewing and accepting your work is overwhelmed and not available. In Scrum, for example, this is why they are so adamant about the importance of the Product Owner role, to ensure that the team always gets fast feedback, focus and vision.

Changing to methodology like Scrum or Kanban will not fix the issue, but will make it hugely obvious and painful and then you will have to hope the "powers that be" choose to help fix it instead of ignoring it.

Answer 2

Regardless of methodology, it sounds like the principle of limiting work in progress would help you a lot.

Rather than undertaking several projects for one requester, would it be possible to focus efforts on the most important one and deliver that before moving onto the next? The requester will be able to focus better on just one project rather than juggling several which in turn should improve their availability for answering questions and reviewing work. Equally, the people who will be using the software will get value quicker as your developers are able to complete work more rapidly.

Also, consider how much work the requester is having to test in one go. If they're having to test several weeks worth of features in one go, it'll be harder for them to find the time to do it. Reducing how much work you are batching up by getting them to test (for example) once a week might help you get feedback sooner and make doing the testing a less onerous task.

Answer 3

This is Business Process Management (BPM). It sounds like your processes are haphazard, non standardized, and out of control. It does not require a ton of sophistication to fix this; you do not need an expensive BPM tool and the smallest changes will bring a ton of value at your level of maturity.

It requires an internal project to be sponsored by your bosses. You need to convince them that an investment of its resources and a few months will bring a lot of value in the work you are doing. It requires you and your team to think about how work flows through your organization now, where you have sticky points and non value-add steps, and then a simple redesign of the flow. You can achieve this redesign with nothing more than sticky notes on a wall to facilitate this analysis.

Answer 4

This scenario is little cumbersome. But following steps can be taken to alleviate the depth of impasse

-> First of make sure you deliver the project in modules. There should not be a single large delivery done at the end of a month or a quarter. If you break them up, ideally you should be delivering once in a week.

-> And, make sure that you appoint one internal candidate as a sub-requester. He should be unbiased and behave and see the product with the vision of a project requester. He should go through all the modules developed and give an instant feedback on it. That way you are reducing the work of Project requester.

-> Once most of the issues and feeback given by sub-requester are solved and are upto the mark, then you will anyway deliver it to the Project Requester. So, you will be assured mostly the product is upto the mark and in the mean time when you sub-requester is verifying the project details, you can start scheduling the appointment with the main Project-requester.

This way you can mitigate the risk.

Answer 5

Could I suggest that you consider bringing in an external help for a few days - and I mean a few, not weeks on end. Don't go for a major consultancy - a small company that has been involved in dealing with companies of similar size to yours will almost certainly be more in tune with your thinking, and possibly at a substantially lower cost too. Get them to suggest a small number of improvements - between two and four will be plenty to start with - that will have a big impact. Make sure they talk to the business people as well as to the IT team.

If the consultant is any good, the improvements will pay for his or her fees in a very short timeframe. Once you see an improvement, get them back in after a few more weeks to see whether there are other significant improvements they can now suggest.

I would love to offer my services, but being based in Scotland when you are in the United States just wouldn't make any sense!

Answer 6

A project management software is a tool. It can't plan and execute the task or your project. The sofware will do what an human being will asks it to do. I still am septical to implement some costly project managment softwares or management systems. But do not get me wrong, we all need a techinical support to do the repetitive tasks more quickly. Organizing projects management in your compagny should be done by the managers of your compagny. But you are saying that they do not seem to have a very good sens of planning and organizationing? I would call a meeting of the main managers involved and put the main problems you are talking about on the table. Maybe put a manager in charge to clean up things. But that person should have all the time he or she needs to do this new job (and not being added to his or her current tasks).