Stop telling me what to do. Tell me what you want to know.
The fastest way to finish a task is to find out it never needed doing.
Opening
Before your next big piece of work, ask one thing: if the answer comes back exactly as you expect, what will you do differently? If the honest answer is “nothing,” you just saved yourself a week. This issue is about finding that question early, instead of after the third failed attempt.
Why do teams pour weeks into work nobody ends up using?
Years ago I ran the machine-learning modeling on a life-science project. The project owner was a seasoned experimental biologist, and he kept handing me spreadsheets to parse. Not one or two. Years of them, filled in by hand by a rotating set of teams over more than six years, long before anyone could point a tool like Claude Code at a messy Excel file. He would tell me exactly what to compute. I built it. I brought back a report. He frowned and asked for a different cut. Then another. Several rounds of this.
The work was real and the parsing was genuinely hard. And not once had I stopped to ask what the parsing was for.
How do you find the one question that makes the work worth doing?
After the third or fourth rejected report I said something slightly rude: “Wait a minute. Stop telling me what you need me to do. Tell me what you want to know.” He told me. I asked why he needed to know it. He told me that too. Then I asked the question that ended the project: “Say the analysis comes back X, or Y, or Z. What do you do differently in each case?”
Nothing. Whatever the spreadsheets said, his next experiments were already decided. The analysis fed a decision that had no branches. For days I had been polishing the input to a function that ignored its input.
The thing everything flowed through was never the spreadsheet. It was the decision sitting downstream of it.
What should you ask before you start the task you were handed?
The task you are handed usually sits a few steps below the decision it is meant to serve. Patch at the task level and you can grind for weeks. Go up to the decision and you sometimes find there is nothing to do at all. Three questions get you there: What decision does this work feed? What are the possible answers? Does any answer actually change the decision?
This is easy to write and hard to live. Plenty of work has no clean decision attached, and “what would you do differently” can be a lazy dodge to avoid something hard. I have used it as both. But every time it saved me a week, it started the same way: someone handed me a task, and I almost just did it.
Loud Camel news
This week I spent on a small version of the same lesson. Loud Camel, a tool that helps researchers get cited and recognized, sends a fair amount of email, and that sending had quietly grown into a scatter of separate paths. Instead of fixing each one, I routed everything through a single point, so now there is one place to see and control what goes out. Same instinct as the biologist story: change the spot everything passes through, not the ten spots downstream of it.
Frequently Asked Question
How do I decide whether a piece of work is worth doing before I start it?
Ask what decision it feeds and whether any plausible result would change that decision. If your next move is the same no matter what you find, the work is optional, however interesting it looks. This is the triage Loud Camel applies to visibility work: instead of a long to-do list, it runs nine high-leverage tactics on a recurring schedule and starts with the one check everything else depends on, whether search engines and AI tools like ChatGPT and Perplexity can find your work at all. If they can’t, nothing downstream matters yet.
This Week on the Blog
How to Increase Research Visibility in 2026 A practical checklist for getting found, not just published.
How academic careers actually work The half of the job nobody teaches in graduate school.
Best AI Tools for Academic Researchers What actually helps, and what just adds noise.
Google Scholar SEO: How Researchers Get Found Small profile fixes that decide whether your work surfaces.
Takeaway This Week
Before your next big task, find the decision it feeds and ask what result would change it. If nothing would, you are already done.