You get an email, you get an email, everyone gets an email!
So what
Our highest-priority review requests lived in loose email with missing details and no tracking. I replaced that with a structured form and automated routing, and it handled an estimated 1,000+ requests.
At a glance
- Role
- Product Manager, at Idaho National Laboratory, a U.S. Department of Energy national lab
- Team
- my team, serving researchers and the reviewing offices
- Timeframe
- during my time at INL (in use through December 2025)
- Status
- Status: Shipped
- Numbers
- 1,000+ requests (est.) · $175K+ in staff time saved (est.)
Maybe, just maybe, our most urgent requests shouldn’t live only in messy email chains.
The problem
When I joined, researchers asked to expedite reviews by email, and most emails were missing something we needed. Every request ate time from three groups. Researchers answered follow-up questions about their own request. Admins chased the missing details, checked the record’s status in the system by hand, then re-typed it all into an email template for the reviewing offices. And reviewers still went back and forth when what reached them was incomplete.
That’s slow. Worse, these were the high-priority items, and with no standard intake and no tracking, one could slip through. Miss one and it’s an information-security risk.
Digging in
I went through past requests to see which details we needed every single time. Then I gathered requirements from my team and from the reviewing offices. The reviewing offices asked for something I hadn’t thought of: text at the top of the form that tells people what to expect before they submit.
Constraints
It had to be quick, cheap and stay out of the core system. Waiting in the shared developer queue wasn’t an option.
What I did
I built a Microsoft Form that asks for everything up front: the record, its current status, the reason for expediting and any deadlines. A Power Automate flow sends each submission straight to my team’s shared inbox in a consistent format. When anyone emailed asking how to expedite, we sent them the link.
The intro text sets expectations: the review queue usually holds 50 to 100 items, and a review takes 10 to 15 business days once the manager review is complete.
Tradeoff: building outside the core system meant it shipped now, with no developer time. The cost is that requests live in a separate tool from the system of record. I took that deal on purpose and planned to close the gap later.
What happened
- Volume: 828 responses as of February 2025, and an estimated 1,000+ requests by the time I left in December 2025.
- Savings: $175K+ in staff time saved (est.). The estimate is salary × time saved for both researchers and reviewers. Researchers spent less time on incomplete emails and follow-ups, and reviewers spent less time chasing missing information.
- Tracking: requests became trackable, which cut down on duplicates and gave us data for forecasting common busy seasons.
Looking back
I scoped the next phase: bringing the request into the core system so it lives next to the record it’s about. If I did it again, I’d capture time-per-request before launch, so the savings figure could be measured instead of estimated.