Open browser tabs are to-do items wearing a disguise. Each one is a deferred intention — an article you will definitely read, a comparison you will definitely finish — and cognitive research on open loops says each deferred intention keeps a small claim on working memory. Seventy-three tabs is seventy-three tiny promises to your future self.
The costs are threefold and measurable: memory (each tab holds real RAM, even discarded ones), attention (task-switching research shows visible reminders of unfinished business tax focus even when unclicked), and retrieval (past a threshold, finding a tab costs more than re-searching it — the point where your tab bar becomes write-only storage).
Enter your tab census. You get a Tab Debt Index, the estimated memory and attention costs, your position relative to the write-only threshold, and a triage protocol that respects the real reason you keep them: closing a tab feels like abandoning an intention. The fix is changing where the intention lives.
The formula
N_tabs- Total open tabs across all windows (log-scaled — tab 80 hurts less than tab 8)
A_age- Age of the oldest guilt-tab
W_windows- Window sprawl — separate browser windows in play
R_retrieval- How often you can actually find the tab you want
H_hygiene- Existing hygiene habits — bookmarking, read-later, weekly closes
How it works, step by step
- Count your open tabs honestly — all windows, all devices you daily-drive.
- Date your oldest “I will get to it” tab. Estimates are fine; shame is not required.
- Rate how often you find a tab versus giving up and opening a new one (the duplicate spiral).
- The model computes debt, memory cost, and your distance past the write-only threshold.
- Run the triage protocol below — it converts tabs back into the intentions they actually are.
Worked examples
The researcher with 212 tabs
212 tabs across 7 windows, fossil tabs over a year old, find-rate 2/10, active duplicates, hygiene 1/10, 16GB RAM. TDI: 99 — Write-only archive, ~17.6GB estimated tab memory (110% of RAM — hello, swap), ~45 min/day attention tax. Prescription: bankruptcy with a dated bookmark folder.
The 22-tab project sprint
22 tabs, 2 windows, oldest a few weeks, find-rate 8/10, no duplicates, hygiene 6/10. TDI: 20 — Working set. High tab count during an active project with working retrieval is not hoarding; the calculator refuses to pathologize a busy week.
How to read your score
Frequently asked questions
How much memory do browser tabs actually use?
Modern browsers average 50–150MB per active tab (we model ~85MB), though they aggressively discard background tabs. The visible cost is RAM; the invisible cost is that discarded tabs reload on click — so a huge tab bar is slow even when memory is managed.
Why is it so hard to close tabs?
Each tab is a stored intention, and closing it feels like formally abandoning that intention — a micro-loss your brain resists (the Zeigarnik effect keeps open tasks mentally active). The fix is not discipline but relocation: bookmarking moves the intention somewhere durable, making the close free.
What is the write-only threshold?
The point where adding tabs continues but retrieving them stops — you open a duplicate rather than find the original. Past this point the tab bar is storage that only accepts deposits, and search-again is objectively faster than look-again.
Is tab bankruptcy safe? What if I need one?
Bookmark-all-into-dated-folder preserves every URL before the close. Practical experience is consistent: people reopen fewer than 5% of bankrupted tabs, which is not a loss — it is the discovery of which intentions were real.
Do tab groups and read-later apps actually help?
Yes, differently: groups help live projects (they preserve working sets), read-later apps help someday-content (they move intentions out of the attention field). The failure mode is using them as prettier warehouses — the hygiene slider asks about closes, not just sorting.
Is my browsing data collected here?
No — you enter counts, not URLs, and everything computes locally in your browser. We will never know about the 14 recipe tabs.