2 Killer KPIs for forecasting in projects

Forecast planning and budget variance
2 Killer KPIs for forecasting in projects

There are, of course, numerous ways to analyse your projects. Budget vs. actual, planning vs. actual, spending variances based on percentage completed, and so on. And all are valuable, but only aimed at looking at the past.

Not that looking at the past is wrong, because this is how you will learn when budgeting and planning future projects. But during your projects, you would mostly like to look ahead. You want to know where you will end up so you can take action in time. In this blog we will address two ‘Killer’ KPIs that will help with: planning variance forecast and budget variance forecast.

Planning variance forecast

This KPI implies the variance between the hours still required to complete the work, and the future hours already in the planning.

  • There are three possible outcomes:
  • 1.  equal to 0: the planning contains the exact hours needed to complete the work
  • 2.  greater than 0: additional hours must be planned to complete the work
  • 3.  less than 0: hours can be removed from the planning, they are not required

In the second case you have information in time to reserve additional hours in the planning. Nothing is as annoying as the tug of war with other project managers over resources, because you realised too late that you still need them. All projects eventually end up there.

The value of the third finding is often underestimated. We often find it quite in order if too many hours have been reserved. We are also rarely rewarded if we complete the work sooner. See also the blog of the 4 time wasters in project planning.

But if you are a service provider, time is money. You don't want to see hours flushed down the drain. You would like to re-assign employees that are no longer needed, on other projects as soon as possible. You should focus strongly on this from a management perspective, because techies continue their tinkering, even if the work is done.

Budget variance forecast

This KPI calculates the difference between the budgeted hours, and the expected result.

  • There are three possible outcomes:
  • 1.  equal to 0: your budget will balance exactly
  • 2.  greater than 0: good job, you will probably have hours to spare
  • 3.  less than 0: you will probably exceed your budget

If you expect to end up positive, you have some leeway. You may decide to transfer hours to another project, assign a more junior employee to gain some experience or you just keep the hours as margin for any setbacks later in the project.

If you will be ending negative, then you have timely information to, for example, schedule a meeting with your customer to adjust the expectations. It can also give rise to changes in the assignment of employees. In any case, you can still do something in time to the adjust the result.

How do you calculate these KPIs?

  • The million-dollar question. Any way, there is a set of principles that you must handle:
  •  Continuously report progress
  •  Hours in the past: actual and otherwise planned hours
  •  Hours in the future: planned hours
  •  An up-to-date view

Continuously report progress

You can have the progress reported by the members of your project team. In the end, they carry out the work and have the best view on the work they still have to do. Or not? In my experience it depends mostly on the person. In addition, you should also take into account the psychological effect that people always like to report progress.

Example: John reported 75% progress in week 1. Then the wheels fell off. It's actually more work than expected. In fact, he is only at 50%. It is very difficult for most people (including John) to report a lower percentage in week 2. They are sticking to 75% or a small percent more.

So you can also consider doing it yourself as a project manager Then you have at least a measure of judging, which might also be somewhat more objective. You will however have to be pretty clued up on things, otherwise you will still have to rely on the members of your project team.

You can measure the progress among other things on the basis of a percentage complete and/or an indication of hours still needed to complete the work. For the best method, read the blog 3 preconditions of project progress reporting .

Should progress really be measured? I believe it should. If you don't, you can still make an estimate. You just take the budgeted hours as a starting point for the hours you require in total. Not very realistic, but it does give an approximation. It is at least better than nothing.

Hours in the past: actual and otherwise planned hours

To be able to make the calculations, you need to know how many hours have been spent in the past. This we can base on the time sheets of the staff. The draw-back is that you often have to wait a whole week before you have the actual hours available. And then you are probably lucky if you have them all. Think for a moment of the ‘usual suspects’ that are always late in submitting their time sheet. That's not actually doable. If you are going to be working with time sheets on a daily basis, you come much closer, but you still have the same problem.

The solution: If actual hours have been submitted in the past, we will of course use this in the calculation. Don't have any? Then we just take the hours in the planning for that employee. In the absence of the actual hours, this is the best approach we can choose. OK, it might differ for an hour or so, but it's still the ‘best guess’ we can take.

An additional advantage of this solution, we can also make this calculation every day of the week. In the absence of actual hours we can always fall back on the planned hours. This allows us to bypass the weekly reporting cycle and acquire a forecast at any time. Even if you are using very exact planning, you would no longer need to keep time for the sake of reporting(!).

Hours in the future: planned hours

In order to calculate the planning variance forecast we need to know how many (future) hours are still in the planning as of today. It is therefore important that you schedule all activities in the resource planning, otherwise we can not make the calculation.

You usually take the current date as reference date. The date of the current day will not be taken into account. The chance of still affecting the course of the current day is not so big. On average you record the status halfway through the day. So half the day is already behind you.

An up-to-date view

An important ingredient is the reported progress. That progress, we as project managers must however assess in the context of current events. If you work with a progress of the same day, then it will most likely be up to date. But what if the latest progress submitted is a week old?

It depends. In a project that has been running for two-years, it will probably still be fairly up to date and therefore useful. If your project only lasts 4 weeks, it is probably outdated and no longer up to date. As project manager, you will have to consider the reliability of the calculated KPIs in the light of current reported progress. Just don't jump to conclusions too soon.

Calculation example

  • And now we will make it more concrete with a calculation example:
  •  A project with one activity
  •  The budgeted hours for this activity are 80
  •  An employee is scheduled for 2 weeks
  •  In week 1, this employee submitted a timesheet with actual hours
  •  In week 2, progress of 30% is reported on Monday
  •  In week 2, we draw up the balance on Tuesday

Schematically, it looks like this:

Calculation examples of forecast variances
  • Based on this information, the following calculations can be made:
  •  Hours in the past = 48
  •  Hours in the future = 24
  •  Planning variance forecast = (70% of 80) – Future hours = 5624 = +32
  •  Budget variance forecast = 80 – 4856 = -24

So 32 hours should therefore be added in the planning to finish the work. There is a negative budget variance expected of 24 hours.

Want more?

For more calculation examples with different scenarios of reporting progress (hours still required and no progress), please refer to the article Forecast Variances on the Timewax support site. You can also watch this training webinar video there.

Thanks to Jurgen Overweg of Evident for his input on the KPIs and the use of his quotes. Questions or comments regarding this blog? Please contact Timewax.

Mark de Jong
Mark is Sales & Marketing Manager at Timewax. He has a background as a project and resource manager with PricewaterhouseCoopers Management Consultants with expertise in the field of Professional Service Automation (PSA)