The “Well, Duh” Episode
- Chris Terrell
- Aug 14
- 3 min read
We started with a laugh, because that’s how most “aha” moments feel in hindsight. You stare at a mess for months, finally pull a thread, and suddenly everyone says, of course. Obviously.
Here’s the story. Years ago, I worked at an e-commerce company with multiple distribution centers. Orders from the West Coast sometimes shipped from the East Coast. We called those “ship-from-wrongs.” Operations kept asking why our top product was always on the list. How could we be “out” of the most popular item?
The answer wasn’t inventory. It was combinations. Customers would buy our top seller alongside some tiny, bespoke SKU stocked only in Pennsylvania. The system had to fulfill the whole order from the location with the bespoke SKU, so the entire order went East. Once we isolated the causal SKU — the item that forced the shipment to the “wrong” coast — and put it in a simple report, the problem became the kind of obvious that makes you grin and groan at the same time. Everyone used the report from then on.
That’s the funny thing about obvious. It’s invisible until you show it.
What makes it appear? Two ingredients: seeing and stakes. People need to see the outcome in a way their brain can’t unsee, and there has to be a consequence attached to the current path. When both show up, adoption stops looking like change management and starts looking like gravity.
We love to think brilliance comes from inventing something totally new. More often, it comes from subtracting confusion. Think Instagram starting as a bourbon-sharing app before users made the real use case undeniable. Think Slack emerging from a side tool that solved a daily annoyance. The “obvious” product wasn’t born; it was uncovered by watching what actually worked.
So how do you design for obvious?
First, hunt for the causal unit. In any ugly outcome, there’s usually one field, step, or SKU making the decision for the system. Your job is to surface it. In sales ops, it might be the one picklist that flips routing logic. In support, it might be the keyword in an email that reopens a ticket. In finance, it might be the single exception rule that shunts a transaction out of the happy path. Build the smallest artifact that exposes that cause: a one-page report, a tile on a dashboard, a column that says “this field decided your fate.” When the decider is visible, the debate gets short.
Second, separate “infinity” work from “one-off” work. QBR decks, daily metrics, month-end closes — these are infinity loops. Treat them like products with owners, SLAs, and guardrails. One-off analysis is exploration; keep it cheap and disposable. Organizations create process debt when a quick one-off quietly gets promoted to an infinity ritual without design, ownership, or data contracts. Label which bucket you’re in before you start.
Third, reduce channel hide-and-seek. “I Slacked it.” “I emailed it.” “It’s in SharePoint.” Pick the canonical home per artifact type and write it down. Then make the rule painfully simple: updates live here, links go everywhere. You won’t get perfection, but you will collapse search time. The measure of obvious is not elegance; it’s “where would a reasonable person look first?”
Fourth, pair a people thinker with a systems thinker. Give a UX-minded operator and a deep technologist joint ownership for a quarter. One asks, “How do humans actually use this?” The other asks, “What does the system actually do?” That pairing finds causal units fast, and they can ship tiny fixes that feel like magic: a keyword filter that stops “thank you” from reopening tickets; a pre-check that flags when a line item will force a ship-from-wrong; a tooltip that explains why a record routed the way it did.
Finally, expect the social curve. Before the reveal, you’ll sound unclear. After the reveal, you’ll sound obvious. That’s not a you problem; that’s how seeing works. Build the demo, show the before-and-after, and let gravity do the persuasion.
Process Debt Truth: Most “genius” fixes aren’t breakthroughs; they’re clear windows installed where the wall used to be.




Comments