Subscribe to Impira's blog
Stay up to date with all things Impira, automation, document processing, and industry best practices.
A long time coming
It's long overdue — you know that now's the best time to get your team's data tasks in order.
When you're the go-to person to tackle data unification and access, your to-do list fills up quickly. You need a framework that saves your team's time and energy. The ideal framework will help you prioritize data initiatives and rethink processes that may be wasting time. A great framework also gives the entire organization visibility into the requests being asked of your team. There's probably a lot more going on than people realize.
So what should you tackle first? What should you automate? Where should you allocate your data budget? These aren't simple questions —they're often overwhelming ones. But while your to-do list has competing priorities, you don't need to succumb to a confusing, multi-tasking spiral.
If you look online, you'll see numerous frameworks for setting quarterly and annual priorities, from OKRs to KANBAN. While the PIE framework isn't specific to data operations, it works well in scenarios where many ideas and needs are thrown at one team in particular. In this article, we'll help you set your to-do list in order with specific actions, then help you think about the ways to introduce automation into your data initiatives.
Introducing the PIE framework
Data teams tackle a diverse array of projects: Unifying multiple data sources, digitizing data from unstructured sources, and data governance. The PIE framework is an acronym for potential, impact, and ease. This type of criteria-based prioritization brings a new layer of transparency to your initiatives, a perfect scenario for teams that get many requests from multiple stakeholders.
The PIE framework, championed by optimization company Wider Funnel, is often associated with A/B testing for website conversion rates but can be applied to data operations or business processes as well. It can help you quantify your potential opportunities and look at ways to optimize for conversion using the data you have at hand.
You'll want to make a list of your big projects on a Google sheet or whiteboard. Then go through each item and score it using the steps below.
In the first part of the framework, you look at the potential for improvement with each initiative.
What part of your data pipeline has the potential to truly improve business outcomes? You can also start by focusing on your worst performers because they have the most potential for improvement.
In this part of the framework, you consider the importance of each data initiative. Which projects have the largest impact on the overall business? Can this initiative change visibility into revenue or reduce the bottom line? Will implementing this change make it easier for a team to do their job? Does it reduce errors and increase trust in reporting?
If you work with business unit leaders to solve data problems within your organization, urge them to quantify the impact that the requested change could make. The chances are that they've already presented this information at some point.
How complex is this task or idea? Your barriers could be technical, organizational, or cultural. Will it be easy to complete in a short amount of time? Will it require procurement of new data processing tools, or maybe hiring a new project manager to keep things moving?
Something that requires little investment and political navigation earns a high score, whereas something that could mean hiring or purchasing would be lower.
Tally up your scores and then divide that score by three to get the PIE value. This is your prioritized list.
From prioritization to automation
Once you've got your final scores, it's time to consider where automation comes into the equation. To help you choose what to automate, think about your data initiatives as follows:
- Put on hold. Some initiatives have low PIE scores but would be very easy to automate. Even though they have a low potential to impact your team and/or your customers, they are quick wins. Put them in your backlog and tackle them once you've completed higher priority work.
- Quick wins. These will be high PIE score initiatives that are also easy to automate and that bring value to your team or organization. Implement these automations right away, as they're sure to improve your workflows.
- Strategic initiatives. These will be high PIE score initiatives with high value for your organization, but they will take significant time and resources to automate. This group might comprise complex automation projects that could temporarily disrupt teams and processes. They should be tackled strategically, with plans laid out for hitting certain milestones over time.
- Rethink. These initiatives will offer low value, and automation will provide questionable impact. These initiatives could be shelved for now or entirely cut out until the company's focus shifts.
As you're going through the PIE scoring process, automation is a natural addendum to your review of your data priorities. It poses a lot of complementary questions, like: "How can we get important work done faster?" or "How can we reduce manual errors?" With both automation and data initiative prioritizing, you're ultimately trying to move your employees' attention toward the most strategic and important KPIs.
The PIE framework is just one way to tackle your data priorities. The important thing is to have something that's rules-focused and brings visibility into the hard work your team does every day. To make your data prioritization even stronger, consider taking some of the following long-term steps:
- Identify a small group of decision-makers to continually assess your data initiative priorities.
- Score automation projects using a similar framework to the one you used to assess your data priorities.
- Share the results of your data prioritization in a central hub where your whole organization can see it.
Bring objectivity and analysis to your must-do list when it comes to data. You don't have to swim between vague goals or become stuck on the question of whether or not to automate certain processes.