3 Steps to Prioritizing Data Requests

As analysts and data scientists, it’s your job to gather data requirements, manage reporting requests, maintain data integrity and obtain confident insights. However, there’s generally one of you supporting a much larger group within your organization. How do you prioritize data requests? How do you turn a long list of data requests, often ranging in difficulty, into a short list of prioritized requests that answer business questions? Do you sacrifice sleep?

Team Chartio (L-R) Jennifer Hudiono and Emily Highstreet

Last Saturday, a few of us from the Customer Success and Marketing teams spent our Saturday at San Francisco’s MeasureCamp, a free data analytics unconference. We spent the day talking to a diverse group of data analysts, scientists and enthusiasts about all things data.

As we spoke with attendees and sat in on round table discussions, we collected insights on how this community of data enthusiasts approach and solve this problem. Here are the three steps to take when prioritizing data requests:

1. Establish the ‘Why’ Behind Requests

No matter the size of your organization, you’re almost always going to work with a finite number of resources. So, the first step in prioritizing data requests is understanding what the business user is trying to achieve with their data. This is absolutely vital to business success (and your sanity). Why?

Well because a simple request like,

“I’d like the number of unique page views for last week’s blog post to date”

Can easily snowball into:

“What was the unique page views for last month’s blog post in the same date range?”

after that

“What was the bounce rate for each in that same date range?”

and

“What was the bounce rate for each in that same date range?”

So on goes a perpetual cycle of questions. You and your data requests start spiraling out of control. Spinning in circles without really getting anywhere, sort of like this:

Starting with the why, or the overall objective, means you can properly deduce urgency, time to completion and relevance in relation to other requests in your queue.

Quick Tip: Getting push back when asking requesters what they are trying to achieve? Remind your teams that data request context lowers the risk of providing bad or incomplete data, which then lowers their risk in making decisions that hurt the business.

2. If you haven’t, implement a ticketing system

A ticketing system is a given when trying to organize or prioritize data requests. When users write out their own requests, ad hoc inquiries naturally decline and inquiries become more thought through. In true data prioritization fashion, there are technologies that will digitize this entire process for efficiency and simplicity. The two most popular tools we heard were Jira and Google Forms (a free option which populates a spreadsheet with submissions automatically and can send alerts).

For optimal prioritization, make sure your tickets include:

• A Category Field that segments requests into type of request (Current Report/Dashboard Enhancement, New Dashboard, One Time Report, Analysis/Insight Only, Data Cleanliness Inquiry, Other).
• Data Description Field communicating the type of statistics or data the user would like to receive and what it should look like.
• Objective Field capturing the bigger accomplishment this data will be used for (this should be required).
• Frequency Field capturing how often this data/information will need to be updated or redispersed.

Based on the above required fields, you can package requests by estimated time to completion and broader objective per team. Next, it will be up to you and the organization to establish a prioritization strategy.

3. Use Executive Weight and Authority

Business objectives and transparency are likely above your pay scale, so don’t stress about it. Let the executives or leads of the business teams decide on prioritization.

Prepare the head of each team with their data request themes categorized by objective type and estimated completion time. Communicate your capacity limits and host a meeting with team leads or executives, bribe everyone with free coffee or Goat Hill Pizza (okay, maybe it’s just an email thread). Next, allow the business teams to prioritize their initiatives, as they are most familiar with the goals and roadmap.

Bonus Points: Some teams are consistently deprioritized due to their function. Come prepared with alternatives such as contractor headcount in exchange for additional budget. Keep in mind, here is your chance to present the business case for additional headcount-use it.

Conclusion

My fellow data teams, you have been honored with data sovereignty over your organizations. We know data is knowledge and knowledge is power. But with great power comes great responsibility. Okay, I’ll stop. But really, prioritizing one project always comes at the expense of deprioritizing another, so it’s imperative we do so transparently and mindfully. As the data heroes in your organizations, I salute you!