5

We currently have a website solution project in which I am involved. We are also outsourcing the development from local software-development companies. We have not selected a vendor because we are at the request for proposal stage.

In the project plan, I need to allocate a timeline for the development of the website solution (which involves more than just a website). How do I go about allocating date/time for this? Best case I need the website to be finished in a month; however, if I put that in the project plan, then it would not be realistic.

I have the website requirements and I thought about placing estimated development times for each requirement however there are too many line items to enter times for.

5
  • Wasn't sure what you meant by 'but it is too many' so I guessed in order to clarify the Question. Feel free to edit if I guessed wrong.
    – Sarov
    Commented Dec 13, 2018 at 14:11
  • 1
    Are you requesting quotes from software-development companies to create the website solution for you, or do you need to come up with a timeline to create a bid on such a project (i.e. you will create it for another company)? Commented Dec 13, 2018 at 14:40
  • You can’t build a schedule solely from targets. You need valid estimates from your engineers or vendors about how long each deliverable will take and what the dependencies are.
    – Todd A. Jacobs
    Commented Dec 13, 2018 at 14:54
  • @BartvanIngenSchenau I think Todd explained what i need to do first. Our project resource usually creates yearly objectives for projects and he gives timelines for the project (i don't know how he comes up with it). So now I am stuck with a 4 month timeline for this website project and trying to create the project scope for it. Can you point out issues with this approach?
    – Raidenlee
    Commented Dec 13, 2018 at 15:08
  • @ToddA.Jacobs You are right but the organization set deliverables without even consulting as you say engineers or vendors for the specific project. I guess they would have looked at stats to determine this.
    – Raidenlee
    Commented Dec 13, 2018 at 15:10

4 Answers 4

4

You have no choice but to get the estimates and planning values--and risks thereof--from the vendors you are planning on using. Then you need to load those external commitments into your master schedule and read the resulting overall duration. Better yet, you would have a worst case duration value, best case, and most likely so you can fully understand what you are facing with your project from a risk perspective.

That duration value from this schedule analysis will most likely differ, and most likely unfavorably, from the four-month target you were given. In this case, you can either lie and tell your sponsors you can get it done in four months and then try to aggressively pursue change requests, or you can tell them the results of your estimating and planning process and put the decision in their hands as to what to do next. The former I would not recommend; however, I can tell you that is done all the time. The latter puts the decision in your sponsor's hands, which is where it belongs. As a PM, you sometimes--and often times--have to tell them bad news.

Their choices range from decreasing scope, extending the targets, or firing you and hiring someone who will give them the former answer above.

2
  • +1 for the best and worst case scenario forecast elaboration! Commented Apr 26, 2019 at 7:41
  • 1
    Thanks David. I went with putting the decision into the sponsor's hands.It worked out pretty well.
    – Raidenlee
    Commented Dec 13, 2019 at 19:35
2

Group the requirements up into Epics (or Feature Groups or however you want to call them...)

Developers estimate for the individual requirements, and then those are aggregated into the Epics - the Epics are then used to build the schedule.

2
  • The issue is that the overall project was set to complete in 4 months without consulting vendors. So i am here trying to place timelines on requirements which i have no idea how long it would take. Great point on allowing the developers to estimate the timeline for individual requirements. I guess i can see how well it fits into the initial timeline.
    – Raidenlee
    Commented Dec 13, 2018 at 15:27
  • @Raidenlee - I have another suggestion, as you can see. Why? Because you have a deadline and the developer's will have to get as much done as possible to fit into that deadline. Commented Dec 13, 2018 at 16:00
2

I would divide it into 4 parts - each one month long, with a milestone at the end of each month:

  1. Definitions - ending with a milestone of agreed upon feature spec
  2. Create initial sites - ending with a milestone of beta sites
  3. Testing beta sites - ending with a milestone of list of fixes for launch
  4. Finalizing sites - ending with a milestone of launch live sites

In reality, once milestone #1 is reached, the coding, design, testing and tweaking happen in parallel - but this gives you a neat way to break it down.

Part #1 also sets the expectations for both sides - here is where the developers get to decide how much they can squeeze into the allocated time.

1
  • 1
    Interesting take @Danny Schoemann , in a world where we want things for yesterday, this shows the acceptance of time. Commented Jul 19, 2019 at 12:49
0

The other answers are great, just one more tip, for safety use more than one estimation method:

Analogy based (like WMFP)

In house expert Score Card

And the contractor estimate (as they know best their team productivity and size)

Then look into any difference between them

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.