At QSM, we have one of the largest industry databases in the world of completed software projects. The data comes from our clients with their permission and this data has been the backbone of our software estimation business for over 35 years. We can see what is reasonable on software development projects as it relates to cost, team size, effort, duration, size, and reliability. Because of our experience we are often asked about risk factors and estimation accuracy early in the project lifecycle. We explain that increased accuracy comes with having historical data and good sizing information.
But what happens on the early estimates when clients don’t have history and detailed sizing information? Can they still generate scope level estimates so they can make good business decisions? The answer is yes. Risk management techniques can be applied and project uncertainty can be calculated so organizations can plan effectively. This is very important because big business decisions are often made early. Decision-makers need to know if they should move forward with a project and they need to know how much time and effort to allocate.
We use SLIM-Estimate, which is a leading estimation tool that leverages the Putnam Model. It generates reliable estimates based on QSM’s time-tested forecasting models and historical data and it also provides scope level estimates when project information is hard to find. It will allow you to see the chance you have of hitting your project goals and it will allow you to factor in your uncertainty.
The process begins with running an initial estimate by entering some basic information. Size information is great to have, but if you don’t have it, you can enter the budgeted effort along with the application type and productivity index. The productivity index is a measure of project efficiency and is calculated within SLIM with industry data or with your own historical data. Once this information is entered, SLIM will then calculate the amount of functionality that can be delivered within an estimated time frame with an estimated team size. This is one example. If you are missing the budgeted effort than you can approach the estimate using other parameters.
This process can be streamlined even further by leveraging QSM’s Feasibility Service. The Feasibility Service is a solution offered within SLIM-WebServices, a web-based version of the SLIM model. These solutions are designed to broaden the estimation capability across the organization, making estimation available to the project stakeholders and contributors. Simply enter your budget, duration goal, and a ballpark size. The project goal estimate is then projected next to an estimate that the QSM Industry Database says is competitive and reasonable. The risk is shown early and can be viewed across the enterprise or accessed by a smaller group.
Once you perform some high level analysis you can then do some contingency planning. Let’s say that you only have a 50% chance of achieving your initial estimate. This initial 50% estimate might be the one that you condition your team to achieving, with a shorter duration and lower budget; though a manager or business analyst might consider adding more buffer. Understandably, you can’t promise the customer an estimate that only has a 50% chance of success. If you have the time and data, you can run another estimate by entering more detailed information as you learn more about your project lifecycle. But if you don’t have that luxury, you can make some decisions quickly by running a Contingency Plan. This allows SLIM-Estimate to take your initial estimate and generate a “safer” one. You can then see the schedule and effort numbers that you have a 90% chance of achieving. SLIM-Estimate will then generate a more conservative estimate, an estimate with buffer built in, one that you can quote a customer.
Once you run these initial estimates you can continue to evaluate the project risk. You might see that you have a 90% chance of staying within a $2 million budget, but only a 70% chance of finishing within $1.5 million. You can then look at some management alternatives quickly in real time. What if you add resources, reduce functionality, or stretch the schedule?
When it’s all said and done, if your estimates end up being 10% different from the actuals, you can leverage historical data to calibrate the SLIM model to adjust for that 10% the next time you estimate a similar project. You can also sanity check your future estimates with your own history and with industry data.
The bottom line here is that important decisions are often made early with only a small amount of information available. It's crucial to have perspective to go along with numbers. Organizations need to be able to apply reliable risk management techniques to their estimation process so they can make the right business decisions early.