You think your project is unique? Think again.

In today’s post, I’m excited to share a method for crafting a more accurate budget forecast for your transformational projects.

Let's begin by debunking a common myth: the uniqueness of your project in coming up with an accurate forecast of it’s cost.

Queue an example…You've been tasked by your CFO to spearhead the implementation of Generative AI within your organization—exciting, right? One of your initial responsibilities is developing the project's budget.

This is a scary proposition to budget for. Why? Typically, IT projects average a cost overrun of about 447%. This means a $10M project might balloon to an actual cost of $45M. That’s a lot of bad budgeting and forecasting. To de-risk that forecast, we'll use a technique known as reference-class forecasting.

This method starts with using experience as an anchor. This is a technique that I teach for project planning in general, but applied to budgeting.

Let’s say you’re planning to renovate your kitchen. You consult two neighbors: one spent $20,000 on a straightforward renovation that took 2 weeks, while the other encountered structural issues and unreliable contractors, culminating in a $40,000 expense and a 6 week project. Averaging these experiences gives you a realistic estimate of $30,000 and a timeline of four weeks. This estimate is grounded in actual outcomes, including both smooth and problematic scenarios, without any optimistic bias.

However, stakeholders often resist this approach, insisting their project is unique and incomparable. But how unique is any project, really?

Let’s apply this to your Generative AI initiative. How can we come up with a reference-based starting point?

Start by collecting data on similar IT projects from colleagues and peer organizations. It’s crucial to acknowledge that while your project might seem distinct, the reality is that unique scenarios are rare. Even if Generative AI projects are scarce, other IT initiatives like ERP implementations or cloud migrations provide valuable reference points. A little logic and creativity can extend these comparisons to your project.

By approaching your project with this mindset, you can set more realistic expectations and improve your forecasting accuracy.