Reduce Scope or Kill the Project?

Question

My goal is to implement business application. I have 10 days to complete, and my estimation says that i need 40 days.

In a team there are two programmers and one programmer-PM.

I assume one functionality takes 3-4 days to implement. So estimation makes us around 6 functionalities possible, and also five is completed. We have to create 20 functionalities.

How to complete complete application in this circumstances? If we assume that not - should we trim user stories for most important? Or it is a dead march?

Answer 1

So, there are a few things important to note.

#1 - Assumptions - You assume one functionality takes 3-4 days to implement. Has this been confirmed with your team? Some tasks might be shorter (or even longer) depending on the work being done. One of the first things you should do is get with you team and get rough estimates (is this 1 hour, 1 day, 2 days) - don't worry about getting precise. All your tasks could be impossible to complete in this time, you will find out.

#2 - Prioritize - You need to prioritize your user stories immediately. Get your team working on whatever code needs to be completed first while you get with someone to prioritize work (unless you are empowered to do it yourself). Work from top priority to lowest priority.

#3 - DEMO/TEST - in such a sort iteration you want to have whoever will be using this testing as you go, not at the end. Take them rough drafts, print-offs, whatever you have to. Make sure they understand and have buy-in with the product. Nothing would be worse than going off for 10 days and finding what you developed didn't meet what was expected and then be way past the due day.

#4 - Pushing your team - this is probably debatable. I believe in pushing my team to work overtime to achieve goals and necessary deadlines. But that said, this only occurs once every few months and I make sure they are compensated for it. If it is not a rare occurrence there is a bigger problem and I wouldn't recommend it.

#5 - Time - if there isn't enough time, then make sure your team is pulling work from top priority to lowest priority. I would also push back against management and try to get a few extra days. Understand why is this the delivery date, and do we have to deliver all of these user stories within that date? Is there potentially a Minimum Viable Product you can deliver and have one or two more iterations.

Answer 2

If you combine this:

My goal is ready to use application for employers and employees. [...] The bad thing is that i have ten days to finish this application.

with this:

We use something based on scrum, making everything in system as needed of working time evidence.

then it's quite clear, that you have just one run. No sprints, no Scrum. Another matter, if you are generally using scrum and want to use it for the project. In this particular case, I would advise against it, and I'll explain it later.

If you combine this:

We have 20 user stories from our project managers and we have to achieve implementation of this goals.

with this:

We all are new to [...] technology and we have to do this in [...]

and add this on top:

I assume one functionality takes 3-4 days to implement. So calculating makes us 6 functionalities, and five is completed.

then it's also quite clear, that you cannot achieve the goal of "completing everything". One can say, that it is highly improbable, however upsetting this may be.

Also, you should avoid assumptions and try to do rough estimates upfront. In such a short time frame, wrong assumptions can be devastating.

How to complete goal in this circumstances, which is completed application? Is it possible? [...] if we assume that not - should we trim user stories for most important?

Your resources and time are limited. Hence your best option is to reduce the scope. While there are a couple of ways of doing that, combining them may give you better results:

1. Define a minimum viable product. Literally cut off everything that is not mission critical. You can come back and revisit those parts later, after your deadline.
2. Get a person, who is capable of prioritizing stories from the end user perspective, and arrange them in order of decreasing value.
3. Start working on the stories from the top to the bottom.

What will happen in the end is you will, likely, not do everything. At the same time, you would have built a valuable "demo", or a part of the final system, which will at least be usable.

What matters in this case is progress tracking. Kanban will be the simplest way to do that, considering the small team and short timeframes. Make a goal of pushing stories through one by one: dev-test-done, as a production line. Preferable testing will start as soon as the first story is completed.

You may also want to sacrifice some quality, thus testing only the main paths and making sure other ones do not damage the data (it is probably OK if they don't fully work anyway).