Atomizing Research: Trend or Trap
Summary
There has been an explosion of interest in atomizing research, where researchers stop writing reports and instead tag individual insights in a database. The theory is that this will unlock your insights, making them findable and reusable. But does it work, and should you be doing the same? At Microsoft we tried it: Five years, 40,000 insights, 20+ research teams, and 17,000 unique users. So what have we learnt? We learnt that a structured insight library can be worth the effort. But we also learnt that atomizing insight can be taken too far. We learnt that context is critical. We learnt that short term efficiency is king, and that the required culture change is hard. And through this journey we’ve begun to discover how to blend atomized insight with your traditional research process.
Key Insights
-
•
Traditional research repositories often become burial grounds for insights, making it hard to find or connect reusable knowledge.
-
•
Shifting to atomic research focuses on individual insights as the fundamental unit rather than entire reports, enabling search, link, and remix.
-
•
Microsoft’s HITS system tracks dozens of thousands of insights connected across products like Windows, Xbox, and HoloLens.
-
•
Creating durable, reusable research insights requires overcoming strong human cognitive biases like hyperbolic discounting and in-group favoritism.
-
•
A culture of continuous curation and managerial priority is essential but challenging to maintain over time.
-
•
Immediate tactical value is key to motivating researchers to contribute to reusable knowledge rather than only producing one-off reports.
-
•
Allowing insights to emerge organically by observing reference patterns reduces complexity compared to rigid categorization of durable vs. raw observations.
-
•
Researchers and product teams benefit greatly from direct linking to specific insights and underlying evidence within their day-to-day workflows.
-
•
Handling conflicting insights involves encouraging researchers to collaborate and rationalize perspectives through syntheses rather than top-down arbitration.
-
•
The just-in-time curation approach folds creation and connection of reusable insights into immediate research tasks, reducing friction and cognitive load.
Notable Quotes
"Repository to me has a connotation of a place where stuff goes to sit and collect dust."
"What if the atomic unit of research wasn’t the reports, but it was the insight within those reports?"
"We wanted a living network of knowledge so yesterday’s insight connects to today’s evidence and tomorrow’s learning."
"Curation is super important, but also super hard."
"You have to push against 200,000 years of human evolution to get researchers to create reusable knowledge."
"Product teams want quick, grab-and-go research results, but reusable insights are more abstract and take longer to digest."
"Providing immediate tactical value is the big one for just-in-time curation."
"Standalone insights destroy the researcher’s ability to storytelling and engagement."
"Durable insights can emerge organically based on usage and referencing patterns in the system."
"The challenge is whether your organization has the energy to overcome cultural and technical barriers driving researchers back to disposable knowledge."
Or choose a question:
More Videos
"People connect with other people, even through screens, so employee recommendations matter digitally too."
Erin WeigelReal-world lessons to improve your conversion rates
June 26, 2024
"Whenever possible, you should be breaking down your work into specific research projects or topics."
Daniel KorczynskiWhy AI Is Bad at Research (and how to make it actually useful)
March 10, 2026
"Sharing meals is a really critical part of building trust which seems kind of strange but breaking bread together makes a difference."
Kim LenoxLeading Distributed Global Teams
May 20, 2019
"We don’t have to restrict ourselves to playing the canary in the coal mine. We can design the harm out of our products before they’ve even taken flight."
Brendan JarvisFraming Tomorrow by Questioning Today
June 8, 2022
"Actors as standardized patients allow safe practice of healthcare scenarios when live environment testing is too risky."
Carol MassaDesigning Health: Integrating Service Design, Technology, and Strategy to Transform Patient and Clinician Experiences
December 3, 2024
"Researchers can overcome reluctance to include others by training them closely and collaborating on synthesis to maintain data quality."
Dr Chloe SharpUsing Evidence and Collaboration for Setting and Defending Priorities
November 29, 2023
"Civic action as proof of concept to get government moving, to get the wheels turning was really interesting."
Sarah Auslander Betsy Ramaccia Gordon RossInsights Panel
November 18, 2022
"Steve just talked about humans being humans in a world full of humans."
Susan Simon-DanielsWar Stories LIVE! Susan Simon-Daniels
March 30, 2020
"Our goal is to automate routine tasks so designers can focus more on research, creativity, and innovation."
Farid SabitovAutomatization for Large Enterprise Teams
January 8, 2024