"I'd hate for us to throw the experience away when we just spent so much time and effort building it."
I heard that sentence a week after the data came in on a project we'd spent six months building. We still hadn't fully processed what the numbers meant yet.
I recognized it immediately. Not because it's inherently wrong, but because I've said versions of it myself over the years.
We set out to build a personalized recap landing page for a big annual conference. The kind of thing that felt like the event actually remembered you were there. Which sessions you attended. Which topics kept pulling you back. A tailored summary of your individual event experience.
Real engineering ambition.
Real design work.
Six months from a team that cared about getting it right.
But somewhere along the way, the API access we needed to do it right fell through.
No individual user data meant we couldn't actually personalize based on your unique experience. So we pivoted to segmentation: figure out which session topics someone attended most, assign them a persona.
You attended a lot of AI sessions? You're an AI Persona. Here's your badge.
In fairness, the badge was accurate. The segmentation worked. The problem was that accurately telling people something they already knew about themselves was not exactly the vision we'd pitched six months earlier.
The original idea kept getting diluted. We shipped it anyway. Given where we were, that was the right call.
The data came back thin.
Not zero. Not embarrassing. Thin enough to tell you something honest about how many people in that audience actually wanted this kind of experience.
That's a finding about appetite, not execution. The API falling through hurt the product, but I don't think it changed the fundamental signal.
I was still sitting with those numbers when the first conversations about next year started, and that's when I heard the sentence.
I understood the instinct completely. Six months of genuine effort from real people. Treating that as sunk cost and moving on feels like dismissing the work (and the people who did it).
But I kept turning it over. And eventually I landed on what was bothering me: the six months are already gone. We can't un-spend them. The question the data is actually answering isn't whether to throw the work away. It's what size of opportunity we were building for.
Thin numbers have a pretty clear answer to that question.
That context changes the next planning conversation significantly. Or at least it should.
It doesn't mean we do nothing next year. It means we need to ask seriously: what does this opportunity actually warrant?
Because there's a version of this project that takes two weeks. A lighter integration. Simpler segmentation. Something that ships fast and still tells you everything the six-month version told you, except you get nearly half a year back to spend somewhere else.
Sometimes it's: right-size the project to match what we know.
The tricky part is that sunk cost reasoning almost never announces itself as sunk cost reasoning.
It announces itself as loyalty. As not wanting to waste good work. As giving a team's effort the respect it deserves.
These are real and reasonable feelings. They belong somewhere in the conversation.
They're just not planning criteria.
What six months of work actually produced was a finding: this opportunity is smaller than we thought going in. The smart move is to let that finding shape the next decision, not work around it because it's uncomfortable.
Have you been in a room where the data pointed one direction and the planning cycle went another?
I'm curious whether right-sizing ever actually won, or whether the room kept reaching for the six-month vision because walking away from the experience felt worse than ignoring the data.
Share your experiences below!
Top comments (0)