Currently, I led complex and tight enough content production for a couple of TV channels. We are pioneer at delivering quality content production, nearly across the middle east.
To have better health checks of the projects, I am planning to apply EVM concepts to proactively manage the integrated efforts.
Now, 90%, I have X amount of episodes for many of the projects with estimated BAC. I believe "Episode Completed" is the most pragmatic way of tracking my % completion and be able to find the Earned Value (EV), to ensure health check and proceed with remaining EVM stuff.
How you normally practically calculate % Completion? I would appreciate mentioning real life experiences only. I have read PMI Practice Standard for EVM many times.
It really depends on the nature of the work being done. For example, if you were asked to edit a book, how would you calculate percentage complete? Would it be:
a) number of pages as a function of total number of pages?
b) number of chapters as a function of total number of chapters?
c) amount of time spent to date as a function of total planned time?
If the duration, production complexity and other factors are roughly the same for each episode, then taking the number of completed episodes as a function of total number of episodes would be a reasonably conservative approach.
Kiron
...
1 reply by Aziz Azzam
May 23, 2022 4:28 AM
Aziz Azzam
...
Kiron,
Yeah, it is a very case-to-case discussion.
I believe the complexity and maybe some other factors can be quite same unless they are part of normal repetitive operations.
It really depends on the nature of the work being done. For example, if you were asked to edit a book, how would you calculate percentage complete? Would it be:
a) number of pages as a function of total number of pages?
b) number of chapters as a function of total number of chapters?
c) amount of time spent to date as a function of total planned time?
If the duration, production complexity and other factors are roughly the same for each episode, then taking the number of completed episodes as a function of total number of episodes would be a reasonably conservative approach.
Kiron
Kiron,
Yeah, it is a very case-to-case discussion.
I believe the complexity and maybe some other factors can be quite same unless they are part of normal repetitive operations. Saving Changes...
First I would say that EVM was developed to find and resolve problems early in a project so if you wait for episodes to complete, you have a trailing indicator that may help you with future episodes, but will mainly reveal issues to late to fix in current work. My recollection is that beyond about 20% complete, EVM quickly loses value to resolve risks rather than just predict how late you will be complete. Since activity completion is too late a trailing indicator in many cases, often credit is taken both at the start and completion of a task (e.g. 50/50).
From my experience in theater, I would break it down smaller than by episode. For instance, you might have discrete tasks for screenplay, story boards, blocking, set construction, lighting, costumes, etc. Those are performed at different times and frequently by different people so you would be better served knowing how each episode is progressing, rather than if opening night comes off on time each time. A complex set for example might need to be started well before the 1st episode where it is required. If you realize you're a month behind on the due date, you're a month behind with no way to fix it.
By developing a WBS for each episode and measuring work started and completed, you have data you can use to manage performance instead of just measuring how many episodes were on-time vs. completed late. Saving Changes...
Thank you. Roughly speaking, we have the same set construction, lighting, costumes etc. nearly for all episodes of particular show/program. I do prepare a production timeline. More less of Episode WBS. Actually, I am not worried about the schedule variance and isn't seeking EVM assistance for scheduling underperformance purposes. Our programs are designed in a way that extremely unlikely circumstances leave room for losing the sight of schedule. At any effort, we meet our on-airing timeline and has done so for long enough. For some programs, I can record 3 episodes in one-day as long as I am environmentally prepared.
The problem I face is the structured and pragmatic determination of earned value, actual costs and more importantly forecasting.
...
1 reply by Keith Novak
May 23, 2022 3:25 PM
Keith Novak
...
In that case I would suggest using some simple six sigma.
EVM will tell you how close your estimates were to your actuals. 6S is all about mean values and variances. By keeping a running total of your mean and the typical variances (SPI/CPI) you have better forecasting data. That can indicate whether trends are changing (you are getting more efficient or episodes cost more) and potential problems like 3 episodes in a row more than 1 standard deviation above the mean.
Thank you. Roughly speaking, we have the same set construction, lighting, costumes etc. nearly for all episodes of particular show/program. I do prepare a production timeline. More less of Episode WBS. Actually, I am not worried about the schedule variance and isn't seeking EVM assistance for scheduling underperformance purposes. Our programs are designed in a way that extremely unlikely circumstances leave room for losing the sight of schedule. At any effort, we meet our on-airing timeline and has done so for long enough. For some programs, I can record 3 episodes in one-day as long as I am environmentally prepared.
The problem I face is the structured and pragmatic determination of earned value, actual costs and more importantly forecasting.
In that case I would suggest using some simple six sigma.
EVM will tell you how close your estimates were to your actuals. 6S is all about mean values and variances. By keeping a running total of your mean and the typical variances (SPI/CPI) you have better forecasting data. That can indicate whether trends are changing (you are getting more efficient or episodes cost more) and potential problems like 3 episodes in a row more than 1 standard deviation above the mean. Saving Changes...
Thanks for the suggestion. I am not that much familiar the way Sigma is used.
Will give it a try and see how far I can make myself understood of the approach.
...
1 reply by Keith Novak
May 24, 2022 6:18 PM
Keith Novak
...
6S can get very involved, but the chart Kiron mentioned is a perfect example of both simple, and very informative information you can monitor.
When you see your data pointing to trends (both averages and outliers), you can dig deeper and look for causes. The causes are really what are really trying to uncover so you can see differences between things like filming on location vs. building sets or using one vendor over another.
Then when you are forecasting, you not only have anecdotal evidence like vintage costumes cost much more than contemporary clothing, you have data to back that claim and a rational basis for estimating how much more.
A simple control/run chart with the completion time (end-to-end) on the Y axis could be used to track whether the episode completion process is in control or not.
Thanks for the suggestion. I am not that much familiar the way Sigma is used.
Will give it a try and see how far I can make myself understood of the approach.
6S can get very involved, but the chart Kiron mentioned is a perfect example of both simple, and very informative information you can monitor.
When you see your data pointing to trends (both averages and outliers), you can dig deeper and look for causes. The causes are really what are really trying to uncover so you can see differences between things like filming on location vs. building sets or using one vendor over another.
Then when you are forecasting, you not only have anecdotal evidence like vintage costumes cost much more than contemporary clothing, you have data to back that claim and a rational basis for estimating how much more. Saving Changes...
