Project Management Central

Please login or join to subscribe to this thread

Historical data for estimating
Network:10



What is the best way to keep track of estimates versus actuals on past projects so you can refer to it for estimating future projects? Is Excel the way to go or are there any tools one can use?
Sort By:
Network:622



I don't believe a tool exists that's specifically designed to track estimates vs actuals. I use MS Project, so I can simply refer to saved schedules for historical data. You can use Excel to do likewise, if you choose.
Network:333



You can do a lot of things with Excel and it has some pretty powerful functionality, but if you want to do more with the data, visualization tools like Tableau are more capable. Your historical database keeps growing, and you may want to look at the data in multiple ways to understand it better. Depending on what you're doing, that can be a lot of recurring work with Excel.
Network:211



Personally, I would use a spreadsheet (either Excel or Google Sheets) because I like the flexibility, and I know how to get it to do what I want (including charts and graphs).

At a minimum, I would track overall project estimate v. actual for each project. If you have individuals providing estimates (rather than teams), then I would also track this data for each individual so that I can learn their multiplier. In the IT industry, most people's estimate is about 2, meaning that it actually takes about twice as long as they estimated. I've only found 2 people in the IT industry that have a multiplier of 1 (ie, their estimates are accurate). Even some of the best software developers are lousy estimators. They are valuable team members, but you need to know their multiplier.

On the other hand, if you do team estimates, planning poker, Scrum development, and other Agile practices, and use velocity rather than traditional estimates, then you sidestep all of these problems, and end up with better results - particularly for projects of high uncertainty, such as software development.
...
1 reply by Zehra Mehdi
Jun 14, 2019 10:10 AM
Zehra Mehdi
...
I like the idea of keeping track of the individual too.
Network:158



Estimation mainly based on WBS & CA. PMP never told any limitation for tools to use. It's based on the choice & tailoring your own feasibility.

You can use Excel, MSP or ant PT for the same.
Network:10



Thank you for the tips!
Network:10



Jun 13, 2019 4:32 PM
Replying to Eric Isom
...
Personally, I would use a spreadsheet (either Excel or Google Sheets) because I like the flexibility, and I know how to get it to do what I want (including charts and graphs).

At a minimum, I would track overall project estimate v. actual for each project. If you have individuals providing estimates (rather than teams), then I would also track this data for each individual so that I can learn their multiplier. In the IT industry, most people's estimate is about 2, meaning that it actually takes about twice as long as they estimated. I've only found 2 people in the IT industry that have a multiplier of 1 (ie, their estimates are accurate). Even some of the best software developers are lousy estimators. They are valuable team members, but you need to know their multiplier.

On the other hand, if you do team estimates, planning poker, Scrum development, and other Agile practices, and use velocity rather than traditional estimates, then you sidestep all of these problems, and end up with better results - particularly for projects of high uncertainty, such as software development.
I like the idea of keeping track of the individual too.
Network:1613



Zehra -

Depending on the type of project, there are commercial estimating tools (e.g. QSM for software development) which allow you to improve their out of the box estimating models by capturing actual cost and schedule performance data.

Other than those, Excel is a popular choice but identifying all the variables which affected performance and capturing those might be tricky...

You could also consider using regression analysis to derive parametric estimation formulas using Excel, Minitab or other statistical analysis tools.
Kiron
Network:103535



I'm not sure how well estimates carry forward even within the same project.

It is amazing how much a team's velocity will change over the course of a project. You cannot compare an early estimate against a later estimate. Both are done with different levels of certainty, risk and assumptions.
Network:333



Stéphane,
We look at a lot of data not so much for just parametric or analogous estimating but to understand the "why". Which group caused it to go late; what step of the process keeps overrunning, was this driven by late change, and those kinds of questions are why we collect a variety of data to enable the data we need to make the right decisions.
...
1 reply by Zehra Mehdi
Jun 17, 2019 5:44 PM
Zehra Mehdi
...
Thanks Stephane, you are correct, but I agree with Keith too, my goal of capturing historical information is to determine what went wrong in the past and what could be done to improve as far as estimating is concerned. It could only be used as a reference for future projects.
Network:10



Jun 14, 2019 1:20 PM
Replying to Keith Novak
...
Stéphane,
We look at a lot of data not so much for just parametric or analogous estimating but to understand the "why". Which group caused it to go late; what step of the process keeps overrunning, was this driven by late change, and those kinds of questions are why we collect a variety of data to enable the data we need to make the right decisions.
Thanks Stephane, you are correct, but I agree with Keith too, my goal of capturing historical information is to determine what went wrong in the past and what could be done to improve as far as estimating is concerned. It could only be used as a reference for future projects.

Please login or join to reply

Content ID:
ADVERTISEMENTS
ADVERTISEMENT

Sponsors

Vendor Events

See all Vendor Events