Using ROI To Prioritize Data Projects

CEO, Portable

How do you calculate the return on investment (ROI) of a data project?

How do you prioritize one initiative against another?

It’s NOT simple.

There are wrong approaches to prioritization…

Not thinking about ROI is WRONG.

Spending all of your time prioritizing work, and not leaving time to solve problems is WRONG.

Keep it simple.

Here’s the framework I use.

1. VALUE - Let executives justify and quantify value

Once you've pulled together a list of high value projects, take the list, identify the champion for each initiative, and ask them to provide a business justification for the project. It could be revenue, cost savings, strategic value, etc.

Focus on the prioritization (i.e. getting stakeholders aligned on which initiatives are more important than others) instead of creating the perfect value metric for each.

Get buy-in across all stakeholders that your prioritization is roughly accurate.

Once you have a prioritization people agree with, find the projects that have quantifiable business value, and add a ballpark number for each.

Then fill in the blanks for the other items. If you have a project that is more valuable than a $1 million project, but less than a $10 million project, ask stakeholders for a ballpark.

It doesn’t need to be perfect, but you’ll want the numbers available to decide what you’ll actually build next.

2. COST - Opportunity cost is your biggest expense

I’m not trying to justify high infrastructure costs. They are a bad idea.

But your TEAM and TIME are the most expensive and most important costs to consider when evaluating a data project.

So how do you quantify the cost of a project?

Work in sprints. One week time blocks where you commit to certain tasks.

For each project, come up with an estimated number of sprints to complete the project.

Do you need to evaluate new tools to solve a problem?

Add the amount of time (in sprints) it would take to evaluate new tools to your estimate for the project.

In almost every scenario, the cost of your team evaluating tools is going to be more expensive that the actual infrastructure.

3. ROI - A simplified model

You now have a ballpark dollar amount for each initiative, and a number of sprints you believe it will take to complete.

Divide the value by the cost, and you can start to see which initiatives provide the best return.

Early on, I’d recommend starting with high ROI tasks that don’t take months or years because you want to show value to stakeholders quickly.

Find projects you can accomplish in 2-4 weeks or faster.

But don’t overthink it.

You don’t want to spend all your time prioritizing, and no time solving problems.

Get to work!