Watch the Market Without Drowning in Tabs.

It is 14:51 on a Wednesday. You have thirty tabs open across two windows — broker, charts, news, e-mail, half a Notion document, three Slack channels, a Calendly, the TFL line status, the back half of a Reddit thread on which you cannot quite remember why you were ever clicking. You are in a Slack call with your team. NVDA prints earnings during the call, the stock moves four per cent in a window of about two minutes, and you find out forty minutes later when somebody messages you about it. The rest of the afternoon, you cannot get the question out of your head: how did I miss that.

You did not miss it because you were not paying attention. You missed it because attention was the wrong tool.

Active versus ambient.

In human-factors research, attention is usually carved into two regimes. Christopher Wickens' multiple-resources model — a workhorse of the field since the 1980s and updated most recently in 2008 — distinguishes between active retrieval of information, which is expensive, and ambient perception of information, which is cheap. Reading a dashboard is active. Glancing at a clock on the wall is ambient. The clock is on the wall whether or not you are looking at it; you can absorb the time, in passing, while continuing whatever you were doing.

Ambient channels are not free. They occupy real estate, they demand a small slice of peripheral vision, they require some discipline in what is allowed onto them. But they cost about an order of magnitude less attention per unit of useful information than active retrieval — provided the channel is well-designed and not flickering.

Most retail traders, most of the day, do not need active market information. They need ambient market information. They need to know that the names they care about are doing roughly what they expected — and to be jolted to active attention only when they are not. The mismatch between that need and the tools we use to meet it is the thing this essay is mostly about.

The browser-tab tax.

The trader described above is not unusual. The typical retail user has somewhere between thirty and sixty tabs open across the working day, divided across charts, news, broker, social and the assorted detritus of modern white-collar work. Most of those tabs are not open because they are actively useful right now; they are open because closing them feels worse than leaving them.

Switching between tabs has a cost. Cognitive-load research into task-switching in general — Rubinstein, Meyer and Evans being a classic reference — puts the round-trip attentional cost of a context switch in the range of hundreds of milliseconds, with a substantial tail. For a browser tab specifically, plausible estimates from productivity studies put the focus-plus-parse-plus-decide window at roughly six hundred to eight hundred milliseconds per switch. Hold the exact figure loosely; the order of magnitude is what matters.

Multiply by two hundred switches a day — a conservative number for an active retail trader bouncing between charts, news and a job — and the total is somewhere between two and three minutes of pure switching cost in the course of a session. That is the easy part of the bill. The harder part is the residue: each switch leaves a trace of the previous context in working memory, and that residue is what makes you forget what you were just about to type, what you were just about to buy, what the sentence you were halfway through actually said.

The cheapest way to reduce the tab-switching tax is to reduce the number of times you switch. The cheapest way to do that is to keep the most frequently checked piece of information out of a tab in the first place.

The slim-bar argument.

Imagine an unusually narrow strip running across the top of your browser, thirty-six pixels tall — slightly less than the height of an average favicon, perhaps a third of the height of a single browser tab. What can you fit in thirty-six pixels?

Quite a lot, actually. A scrolling row of thirty US-listed symbols, each with its ticker, last price, percent change, and a sixty-tick sparkline. A small badge for which session we are in — pre-market, regular, or after-hours. A subtle colour cue for direction. That is the entire density budget. It is enough to know, at a glance, whether your watchlist is calm or interesting, without ever moving your eyes off the page you are actually reading.

What you cannot fit, on purpose, is a chart, a headline body, a research note, a stat block. Those things live one click away — in our case, in a panel that slides down underneath the symbol you tapped. The constraint is the point. The bar is the ambient channel; the panel is the active one. You only pay the active-attention cost when the ambient channel told you something worth paying for.

We have written separately about what to actually put on a working watchlist — picking thirty symbols out of the universe is a meaningful constraint and worth its own post — and about how the ribbon shape compares to the other options in our survey of Chrome ticker extensions for 2026. The short version is that thirty symbols is the upper edge of usefully ambient. Beyond that, the bar starts to feel like a screening tool, and screening is an active task.

What an always-on bar should never do.

Ambient channels are fragile. They earn their place by behaving themselves. There are three things a top-of-tab ticker bar must not do.

1 · Never block site headers

The most common failure mode of overlay extensions is that they cover the top sliver of the page they are running on. That breaks the back-button affordance on Twitter, the compose menu on Gmail, the file tree on GitHub. RIBN instead offsets documentElement by exactly thirty-six pixels — the bar's height — so the site keeps its full normal layout and slides down under the ribbon. Sites we have explicitly tested include X, Gmail, YouTube, GitHub, the New York Times, Linear, Notion and Google Docs. The offset is reapplied through SPA navigations via a MutationObserver, because Twitter and GitHub do not full-reload between routes.

2 · Never break SPA navigation

Single-page apps make life harder for content scripts. They rewrite the DOM without tearing the document down, so an extension that injects itself once and forgets is quietly broken on every subsequent route. The defensive fix is the same MutationObserver, watching the top eighty pixels for fixed or sticky headers and re-applying the offset whenever the router replaces them. This is the boring kind of correctness; you only notice it when it fails.

3 · Never follow you into a Zoom call

An ambient channel that you cannot turn off is no longer ambient — it is an imposition. Per-domain hide is the feature that makes the rest viable: Cmd+Shift+H toggles the bar on the current site, a blocklist in settings keeps it permanently hidden on the sites you choose, and Do-Not-Disturb mode silences alerts during focus blocks. The bar shows up where it is wanted and stays out of the way everywhere else.

The flow on a working day.

The texture of using an ambient channel is hard to convey in feature copy and easier to describe as a small chronology. So.

09:25 ET. You open the laptop. The bar is already there — the service worker spun up the moment the browser process did, the WebSocket reconnected, the last prints from the previous session are visible. You open your email. You do not open a chart.

09:30 ET. The market opens. You do not do anything. The bar does the work; symbols turn green or red in your peripheral vision while you reply to messages. Within a minute or two you have a felt sense, without having read anything, of whether the day is calm.

11:14 ET. The TSLA sparkline turns red on a real move — not the noisy one-tick wiggle, the kind that shifts the whole curve. You glance up, click TSLA, the panel slides down beneath it. The chart is right there; the latest five headlines are right there; one of them tells you a recall has been announced. You read for thirty seconds, close the panel, get back to whatever you were doing. Total attentional cost: under a minute. No tabs opened. No context lost.

14:02 ET. You have a meeting. You hit Cmd+Shift+H and the bar disappears for the hour. The meeting ends; the bar comes back; the world has not ended in your absence.

16:00 ET. The close. The bar settles. You did not stare at a chart all day. You did not accumulate twenty new tabs. You knew, more or less in real time, what your watchlist was doing.

The argument, compressed.

Active information requires a tab. A tab requires a switch. A switch costs roughly six to eight hundred milliseconds plus a residue of working-memory disruption that is hard to quantify but easy to feel. Two hundred switches a day is normal. The cumulative bill is large enough to want to reduce. The cheapest way to reduce it is to move the highest-frequency check — your watchlist — out of the tab system altogether and into a persistent ambient channel.

A thirty-six-pixel ribbon at the top of the browser is the smallest channel that fits the watchlist comfortably, the largest one that does not interfere with the page underneath, and the only one that is genuinely everywhere — Gmail, GitHub, your broker, your CMS, the Reddit thread you do not remember opening. The sister piece on stock alerts that do not expire argues the same thing for the deferred case: the bar handles ambient awareness while the alerting backend handles the events that demand active attention. Together, they are most of what a working trader needs.

FAQ