Topics
December 15, 2025
December 15, 2025
Picture this scenario: you’ve found a research platform that is an all-around better fit. Perhaps it offers a wider range of methods, higher-quality participants, or greater accessibility across your team (or all of the above), but there’s one problem…
You have years of data on a different platform.
The task of sorting through endless materials and determining what to move to a new tool seems impossible. So, even though the new option is a better fit, how can you let go of that data?
The answer is frightening but true: you don’t need it all. Even if you aren’t switching platforms but want to create a new research process, cleaning up old data is an important practice.
Below, I’m going to help you break down what data is worth saving and what you need to “thank for its service” and let go of (shoutout Marie Kondo) because old data should not prevent you from elevating your research practice.
While it does feel important to keep everything, research data (and the insights that come from it) have a shelf life.
Before you jump into organizing, let’s start by clarifying a few key terms:
I also want to call out research repositories: a centralized, searchable archive for research insights/findings.
Repositories differ from company to company, but in essence it’s a way to store, comb through, and keep research findings alive.
While the shelf life across data and insights differs, they all do spoil eventually.
Understanding how long these materials are relevant is the key to deciding what to take with you and what to leave behind. There is no science here though, sorry to break it to you, I have no formula to share, but rather some principles to help guide you and determine your own formula or philosophy.
Research data often becomes stale very quickly.
A minimalist like myself might say that this data becomes stale almost as soon as the project “wraps” and all artifacts are created and stored, insights gleaned, shared out, stored, and repository records updated.
I can say without a doubt that I have never returned to a random interview or watched an old usability recording from years ago to listen in “just in case.” Now, of course, I have checked back into old research reports and readouts to remind myself of a study’s content or even pull a relevant quote or video clip.
Key takeaway: Most of your old, raw research data can be left behind.
Research artifacts grow stale over time because products evolve, organizations shift, industries advance, and human behaviors change.
If you’ve just completed a study about a new navigation pattern in your tool, implemented that pattern, and now are getting post-release feedback—some of those early concept tests might not really matter anymore, as you likely have new data coming in that is more relevant and accurate.
“But what if the release doesn’t go well? What if we need to go back to the drawing board? Will any of those artifacts from early concepts be relevant again?” Honestly, probably not, and goodness me, I hope we don’t find ourselves in these situations frequently.
Key takeaway: It’s safe to assume that most of your years-old research artifacts are out of date, you have newer/more relevant data rolling in, and don’t need to be carried over.
"Without regular maintenance, it can quickly get unwieldy and then lose its efficacy. Even worse, it could lead folks down the “wrong” path when insights become irrelevant due to societal or cultural changes."
Research insights have the longest shelf life because they can tell us things about our users, our industries, and our tools that extend beyond the simple objective of a singular study.
But not all insights are like this. Some insights matter only in the context of that study and might become outdated as quickly as the UI or features do.
Sometimes these insights can help us determine future curiosities and considerations. That is why I feel the insight to repository pipeline is much clearer, as these insights can stick around for the test of time.
Insights can exist outside of decks or readouts; they are transmutable (or should be) to fit into the context in which they will have the most impact.
Creating a process or standard for cleaning and organizing repositories is a science; that’s why we have library scientists and research ops experts in our field.
Just as dictionaries change (I mean, “skibidi” was just added), your repository could hold a decade's worth of insights, but still require updates and adjustments.
Without regular maintenance, it can quickly get unwieldy and then lose its efficacy. Even worse, it could lead folks down the “wrong” path when insights become irrelevant due to societal or cultural changes.
Take a look at AI, for example. Many of our insights pre-AI boom have to be seen through a new lens now. Not to mention the changing landscape of your tool, its technology, interfaces, capabilities, etc.
If parts of the tool simply won’t exist anymore, or the strategy has shifted significantly away from addressing certain jobs to be done (JTBD), those insights won’t serve the new reality.
Even though it may seem daunting to switch tools and “lose” a good portion of that legacy data, a cleanup might actually be overdue. There are two schools of thought here (but please consult your local library scientist or research ops specialist):
In true Marie Kondo style, put all of your clothes on the floor and finish the task before you can move on with your life. This might not be possible depending on how many years and how robust your repository has become.
Our team most often takes this approach, constantly culling items and keeping them up to date. Something I like about this method is that it means we touch these insights more frequently—they don’t pile up and get covered in dust because we’re shuffling them with consistency.
If you’re in doubt, don’t “throw it away” when it comes to insights. Or maybe start with an “archive” tag that clears things out of your main views.
Don’t let years (or decades!) of old research data prevent you from pivoting tools or creating a new research process.
It’s time to let go of a good chunk of old research data and artifacts, and start thinking about your philosophy on insights expiration criteria.
I’m not saying start chucking Google Drive folders in the trash willy nilly, but also, it might feel good to get rid of your camcorder recordings from the field in 2017 (I’m speaking from experience, I assure you).
And while I’m not a data security admin or on your legal team, there might even be policies at play that you need to check in about data deletion and storage! We’ve all been there (see 2017 footage above), so now is the time to check in on your company policies here as well.
If you are a maximalist reading this or the sentimental type, I mean you no harm! If you get peace from having all the data you’ve ever collected ever, I see you, and I want you to ask yourself: Am I keeping this data out of fear? Fear of a future impending question that will likely never come?
If so, maybe today is the day to confront that fear and rid yourself of some extra cognitive load and, honestly, storage constraints.
If in five years, you really did need that old footage that you burned and it would’ve saved the planet or changed the world, I promise you can call me up—I’ll buy you a drink and cheers the data’s demise.
P.S. You can also throw away those old iPhone boxes. I promise you don’t need them.
