How to Implement Agile Earned Value Management (EVM)

Agile brings undeniable benefits to project management. But at the same time, it raises new challenges for those who manage software projects on a regular basis. Because Agile embraces change so openly, it increases the difficulties of controlling resources in changing conditions. At the same time, it complicates cost calculations — if conditions keep changing, how can costs be calculated accurately?

These key challenges may have a direct impact on the expectations of business owners, who normally expect projects to be completed on time and within the available budget. Apart from these eternal expectations, business owners may also want to know how much more time and money the project will require and whether they should allow for a longer development time to achieve the initial project objectives, or scale down the features of the software to finish the project before the deadline.

Benefits of EVM

The Scrum framework on which Agile software development is based isn’t very clear about cost control. Metrics such as story points, velocity, feature count, or sprint burndown don’t add up to create a clear picture as far as cost efficiency is concerned. For companies, this can prove to be a major gap in an otherwise highly effective software development strategy.

The Earned Value Management approach for Agile comes in to fill this gap. It comes with some important advantages.

•    Provides comprehensive metrics for cost, performance, and schedule, allowing you to calculate cost and schedule accurately, make forecasts, and comparisons for planed, earned, and consumed value
•    Ties in more abstract valuations like size and complexity with easily observable time and cost metrics
•    Reflects scope changes using the Scope Floor variable

Earned Value Management (EVM) is a reliable technique for measuring project performance and progress objectively, using the latest data available. EVM is effective at spotting project performance problems early on, giving you enough time to address them, so that the project is completed on time at the desired quality level. Its power lies in its integrated approach to measuring scope, time, and costs — the project management triangle. EVM has become increasingly popular in recent years, as new studies have found that the principles it uses are reliable predictors of project success.

Another way to understand the benefits of EVM is to consider a project that does not include EVM and one that does.

Project #1:
A project and its spend have been planned in advance, with a detailed spend plan for every element and stage of the project. As the project advances, the cumulative cost is tracked and compared to the predicted cost. During certain weeks the project may appear as either over budgeted or under budgeted. But what the project manager doesn’t know is how much actual work has been done during each of these stages.

Project #2:
Same as project one but with EVM, which quantifies accomplished work. With EVM, the manager is able to identify which element of the task has been done. He can then apply the principles of EVM to each of these to determine whether he is on budget or under budget based on the amount of work done.

Understanding Agile EVM

Earned Value Management is an old concept that has been around for over 50 years. Only in the last 10 years or so that it’s made its way into Scrum. Its resurrection isn’t an accident. The great thing about Agile EVM is that it starts from using the latest Story Points estimate, rather than initial estimates before the project took off. The result is a much more accurate prediction that reflects the state of things.

Agile EVM comes with a number of key concepts.

•    Expected Percent Complete (EPC) – Amount of value planned to be delivered over a certain period
•    Actual Percent Complete (APC) – Amount of value actually delivered over the period in question
•    Planned Value (PV) – Budgeted cost for the work scheduled to be completed up to a given point
•    Earned Value (EV) – Budgeted cost for the work actually completed during a given time period
•    Actual Cost (AC) – Budgeted costs in PV for release
•    Budget at Complete (BAC) – Initial budget for the release
•    Schedule Variance (SV) – Monetary difference between the planned and actual time to complete a certain workload
•    Cost Variance (CV) – Monetary difference of value incurred and cost to produce a certain piece of work over a certain amount of time
•    Scheduled performance and cost performance indexes (SPI) and (CPI)
•    Estimate to complete (EAC) / Estimate at complete (ETC)
•    Variance at complete (VAC)
•    To complete performance index (TCPI)

Don’t worry — you don’t have to memorize these to make the most of Agile EVM.

How to Implement Agile EVM

EVM includes a wealth of indicators, making it seem daunting at first. In the beginning, though, it’s better to focus on its key features to keep things simple. To implement EVM you need, in addition to a project plan, a valuation / budgeted cost of work scheduled (PV or BCWS), and metrics that quantify the work being done, that is, earned value (EV) or budgeted cost of work performed (BCWP).

Initial Preparations for Setting up Agile EVM

1.    Gather project work as work items (Story Points)
2.    Manage all Story Points in a ordered list (Product Backlog)
3.    Determine the Planned Value (PV) of the project by estimating project size compared to other items
4.    Establish a definition of “project completion,” by listing the criteria a project should meet before it’s being listed as complete
5.    Give Story Points for each completed project based on its complexity

The easiest way to implement Agile EVM is by using a spreadsheet. With a spreadsheet you can automate calculations. But at the same time, you have full control over all numbers, allowing for fast and easy adjustments on all numbers. You’ll start from basic estimates for scope, time, and cost. Here are the input metrics you can use.

Release Baseline
•    Estimate scope in planned release points (PRP)
•    Time in planned sprints (PS)
•    Cost in budget at complete (BAC)

Sprint Baseline
•    Scope in story points (SP)
•    Cost in planned value (PV)
•    Time in starting date (SD) and length (L)

After each sprint, collect relevant data using the metrics for number of sprints completed, number of story points, actual cost.

Once you have them, you can calculate the following:

•    Scope = actual percent complete (APC) against expected percent complete (EPC) to determine how much of the scope has been done
•    Cost = earned value (EV) against actual cost (AC) to determine the monetized value earned
•    Time (number of sprints done) = cumulative per iteration

So what can you actually do with these metrics?
Compare guesses for key considerations such as time, cost, and scope. Determine the efficiency of your planning and performing. Change the baseline on project estimates to reach accurate values.

Find How to Effectively You Are Spending Money and Time

Cost Performance Index (CPI) = Earned Value divided by Actual Cost
•    You’re on budget if CPI = 1
•    You’re under budget if CPI = >1
•    You’re over budget if CPI = <1

Schedule Performance Index (SPI) = Earned Value divided by Planned Value
•    You’re on schedule if SPI = 1
•    You’re ahead of schedule if SPI = <1
•    You’re behind schedule if SPI > 1

Scope Change and What to Do About It

Whether it’s due to increase story points or team size changes, scope changes are ultimately unavoidable for Agile projects. But you can factor them into your calculations so that they won’t take you unawares. Factoring these into your Agile management strategy and calculating them helps to improve your project management process.

It’s important to make the distinction between changes and adjustments. Adjustments don’t happen by chance and are not beyond your control. For example, a change is caused by faulty features or product defects that alter the pace at which the project is being developed. An adjustment, on the other hand, is something that you do manually when the developer leaves the team or if a new feature is added.

Agile EVM has a variable known as Scope Floor (SF) which reflects the impact of the scope change on time and cost. Determining the Scope Floor isn’t that hard. In Agile EVM, scope is determined by Planned Release Points (PRP), which in turn is calculated based on Story Points added (SP+) and Story Points removed (SP-).

Using Agile EVM to Make Cost and Schedule Forecasts

The power of Agile EVM lies in its ability to accurately predict future and past performance for near-terms. This comes in handy in a number of situations. You can use it to predict the total costs of a project after a scope is changed or the cost that the project will require from the present date until the planned end of the project, as well as discover cost discrepancies between predicted costs for a project and actual costs based on the performance of the project so far. You can also determine the efficiency of a given project. Here’s how you can do all that.

•    Not sure how much more money is needed to complete the project from the current day until the deadline, after a scope change?
Estimate at Completion (EAC) = ACcum + BAC – EVcum

•    Worried that there is a major difference between the planned costs for a project and the actual costs the project will require?
Variance at Complete (VAC) = BAC – EAC

•    Not sure whether to make changes to a project? You can use Agile EVM to predict its projected efficiency/
To-Complete Performance Index (TCPI) = (BAC-EVcum) / (BAC-ACcum)

With all these forecasts, you will be able to make more informed decisions about ongoing Agile projects. You can decide when to release the product, at what cost, and with what features.
Agile brings undeniable benefits to project management. But at the same time, it raises new challenges for those who manage software projects on a regular basis.
Possible Limitations of EVM

As most management methodologies, EVM comes with a number of possible limitations. These may become apparent only in certain contexts. It’s important to be aware of them in order to mitigate any risks they may give rise to.

•    EVM was not designed for projects that require continuous work, or non-discrete effort (level of effort = LOE). Projects that contain both discrete and non-discrete effort should not be measured using EVM standards, or the results will be inaccurate. For Agile EVM to work, it’s crucial to ensure you apply it only to projects that require discrete effort.
•    EVM relies heavily on accurate cost data and accurate project accounting. Without this data, it may not be possible to get accurate predictions or even implement EVM at all. The problem is that many smaller companies either don’t have this data available, or else they have to hire a professional to produce it. In such scenarios, companies can limit themselves to implementing EVM only for larger, more important projects.
•    EVM may also be vulnerable to data timing errors. If the data is not up to date and in sync with the actual progress in development for the project — whether as a result of slow reporting or misreporting — EVM calculations will be off the mark. By the time the analysis is delivered, the results may be behind the actual results and thus unreliable.

The Bottom Line

While implementing Agile EVM is not without its challenges, it can be used even by small software companies as a means to better manage the development of their products, make better predictions about key aspects of their projects, and ultimately, create better products.

On the plus side, Agile EVM implementation is not costly and does not require any massive changes in the way a company manages projects. There’s no need for costly software or hardware. That said, Agile EVM does rely on accurate data and good math. Mistakes can creep into Agile EVM, which in turn can lead to poor project management decisions that defeat the whole purpose of its implementation.

Ultimately, Agile EVM is a reliable methodology for businesses that want to make the most of Agile while filling in the gap that the former leaves in terms of cost control. EVM complements Agile, making it a more reliable and ultimately more accurate way to manage projects and release successful products.