Research starts with a question and ends with seventeen tabs. Each new tab feels like progress. Most of them are rabbit holes that never feed the decision. Cap research tabs at five before any rabbit hole so curiosity stays bounded by a hard count.
Five is arbitrary on purpose. The number is small enough to force triage. When the sixth idea appears, something must close or wait. That friction is the feature.
Open a five-slot tray before the first search
Create five named slots: question, primary source, opposing view, numbers, decision note. Leave them empty. Every new page must claim a slot or it does not open. Untitled curiosity does not get a seat.
Write kill criteria first when the topic is shiny, using write kill criteria before you research a shiny idea, so research has a stop rule before tabs multiply.
Fill slots with one page each, then stop
Each slot gets one strong page, not a pile. If a better page appears, replace the weaker one. Replacement beats accumulation. When all five are filled, you research by rewriting the decision note, not by opening slot six.
Park rabbit holes on a later list
Interesting side questions go on a dated list with no links yet. Links invite return browsing. Titles alone are enough. Schedule a separate curiosity window if needed. Mid-research curiosity is how five becomes fifty.
The same door used when you mute curiosity tabs before they eat your next hour applies here. Mute first, browse later.
Argue the opposite with one of the five
Reserve one slot for a hard opposite. If every tab agrees with your hunch, the tray is biased. Spend five timed minutes on the opposite case like argue the opposite case hard for five timed minutes before you call the research done.
Close with one cheap test, not more tabs
Turn the decision note into one falsifiable check. Fifteen minutes. If the check needs data you do not have, redesign the check once. Do not reopen the browser farm. That pattern matches split a stuck hard problem into one next cheap test.
When decisions still spin after the five tabs, put a clock on the choice with timebox looping decisions so choices stop spinning instead of hunting a sixth source for comfort.
If a teammate drops ten links into chat, triage them into your five slots or the later list. Do not open all ten out of politeness. Politeness is how shared research becomes shared wandering.
Print or screenshot the five-slot tray at the start of hard topics. Visual emptiness makes the sixth tab feel like a rule break instead of a harmless click. Rules you can see beat rules you remember vaguely.
End each research block by writing the decision in one sentence even if the decision is to wait. Waiting without a sentence becomes another rabbit hole tomorrow under a new title.
FAQ
Why five tabs?
Small enough to force triage. Large enough for question, source, opposite, numbers, and a decision note.
What if I need PDFs and tools too?
Count each as a slot. Tools are not free.
Can I bookmark extras?
Yes, into a dated later list without opening them now.
What if a source spawns subpages?
Stay on the parent unless a subpage replaces a weaker slot.
Does this slow deep research?
It slows wandering. Directed research gets faster because slots stay honest.
How do I handle team shared research?
Share the five-slot tray. Extra links go to the later list with owners.
