Can Chrome open a specific website every time a new tab is opened, without an extension?

No. Chrome's settings can point the startup screen at a specific page or set of pages, and the Home button at a custom page, but the new tab page itself is not addressable from Settings. Changing it to a web address requires an extension that declares the `newtab` override, which is why that category of extension exists at all.

Why does a new tab extension stop working in an Incognito window?

Because Chrome does not allow it. The developer documentation states that in incognito windows, extensions cannot override New Tab pages, so the built-in page appears instead. Anything stored on the replacement page, such as notes or a link list, is unreachable from a private window for the same reason.

Why does typing straight after opening a new tab go to the address bar instead of the extension's search box?

That is Chrome's behaviour, not a bug in the extension. Chrome tells extension developers that new tabs give keyboard focus to the address bar first, so the box drawn on the page has to be clicked before it receives anything. For people who open a tab and type in one motion, this alone can make a dashboard slower than the default page.

Do new tab extensions make Chrome slower?

The page itself is one more thing to load every time a tab opens, which is why Chrome's guidance to developers is to keep override pages quick and small. The larger effect is usually elsewhere: extensions with broad site access run on every page, not only the new tab. Chrome's own speed guidance lists turning off unwanted extensions alongside closing unused tabs and reviewing the **Preload pages** setting under Settings, then Performance.

Is it safer to use Chrome's own page with a custom background instead of an extension?

It avoids the permission question entirely, which is the main risk in this category. The trade-off is that the built-in page cannot be pointed at a web address and cannot show anything Chrome does not already show. For anyone who mainly wanted a calmer screen and their own shortcuts, that trade is usually worth taking.

What replaces the new tab page for someone running several web apps all day?

Installing each service as an app gives it a window of its own, so it never has to be launched from a new tab. Where the same service is needed under two accounts, a single browser profile cannot hold both signed in at once, so the choice is between separate Chrome profiles and a browser that keeps each service in its own container.

A new tab for Chrome: making the first screen useful again

Cmd + T gets pressed more often than almost any other shortcut, and what appears afterwards is a search box, a grid of thumbnails Chrome chose, and sometimes a card suggesting something to read. For a lot of people that screen is dead weight. The tab was opened on the way to somewhere specific, and the page in the middle is not where they were going.

Searching for a new tab for Chrome usually lands in the Chrome Web Store, in a category full of extensions that all promise a dashboard. That is one answer. It is not the only one, and it is not free. What follows separates four things that usually get mixed together: what Chrome already changes on its own, what an extension has to do in order to replace that page, what replacing it costs, and the case where the new tab page is the wrong thing to be fixing.

What Chrome already changes without an extension

More is adjustable here than most people realise, and none of it requires installing anything.

The entry point is the Customize Chrome button at the bottom right of the new tab page. Under Shortcuts, the grid of tiles can be switched between My shortcuts, where the entries are chosen by hand, and Most visited sites, where Chrome picks them. Individual sites can be edited or removed. Under Appearance, Change theme offers a gallery or Upload an image, and Chrome then suggests colours that pair with whatever image was chosen. The same section holds the Light, Dark and Device setting. Customize Cards turns the content cards on and off individually, and Footer has a Show footer on New Tab page toggle.

What none of that does is point a new tab at a web address. That distinction trips people up because Chrome does have settings that point at addresses, and they apply somewhere else.

The startup setting offers three choices: Open the New Tab page, Continue where you left off, or Open a specific page or set of pages. The third one is the setting people find when they go looking, and it only applies to the pages Chrome opens when it launches. The Home button is a separate control again, and Chrome documents exactly two options for it: the New Tab page or a custom page.

So a Home button can be a specific site. The startup screen can be a specific site. A new tab cannot, not from Settings. That gap is the entire reason the extension category exists.

What an extension has to do to replace the page

Replacing the new tab page is a documented capability with documented limits, and reading them first removes most of the surprises.

An extension declares the replacement in its manifest under the key chrome_url_overrides, and the value names a page inside the extension. Three Chrome pages can be overridden this way: bookmarks, history and newtab. Chrome's own developer documentation states that each extension can only override one page, so an extension cannot quietly take over the history page as well as the new tab.

Two limits matter more in daily use than either of those.

The first is incognito. Chrome states plainly that in incognito windows, extensions cannot override New Tab pages. A replacement page that holds notes, links or a reading queue simply is not there in a private window, and the built-in page comes back instead.

The second is where the keyboard is pointing. Chrome's guidance to extension developers is that new tabs give keyboard focus to the address bar first, and that developers should not assume focus starts somewhere on the page. That single sentence explains a complaint that comes up constantly: a dashboard extension draws a large, inviting search box, and typing straight after opening the tab goes to the address bar instead. The box has to be clicked first. For anyone whose habit is to open a tab and type immediately, a page built around its own search box is slower than the page it replaced, not faster.

Chrome also tells developers to keep the override page quick and small, on the grounds that users expect built-in browser pages to open instantly. That is a reasonable thing to check before committing to an extension: open a new tab ten times and watch whether anything flashes, reflows or arrives late.

The original page does not disappear, by the way. Typing chrome://newtab reaches it.

The permission question, answered by what the page needs to draw itself

Extension permissions are where this decision stops being about taste.

Chrome's documented permission for site access reads "Allow this extension to read and change all your data on websites you visit", and it can be narrowed to three levels: When you select the extension, On a specific site, or On all sites. Chrome's own advice alongside that setting is to make sure only extensions that are trusted get approved.

The useful way to apply this to a new tab replacement is to ask what the page needs in order to render. A page that shows a set of links, a note field and a background image needs storage and nothing else. A page that shows weather, unread counts, or a feed needs network access to fetch them. A page that builds its own list of frequently used sites needs read access to history or bookmarks, which is a much larger ask than it sounds, because a list of visited addresses is a detailed record of a working day.

An extension asking for permission to read and change data on all sites in order to draw a clock is worth closing. An extension asking for storage in order to hold a list of links is asking for what it needs. The gap between those two is the whole judgement.

There is a related point about the built-in page that gets overlooked. Most visited sites is generated from browsing, which means the default new tab page is a summary of recent activity rendered at full size. Anyone who presents, screen shares or works in an open office is showing that grid to the room every time a tab opens. Switching to My shortcuts is a two click fix, and it is a better reason to change the page than aesthetics.

Four shapes of new tab, and what each one trades away

These are genuinely different tools rather than better and worse versions of one tool, because they solve different halves of the problem.

Approach	What it gives	What it costs
Chrome's page, customised	Shortcuts by hand, own background, cards off, no install	Still cannot be pointed at a web address
A blank or minimal override	Instant, nothing to look at, nothing to load	No launcher at all, so the address bar does all the work
A dashboard extension	Links, notes, weather, a feed, one place to land	Permissions to review, absent in incognito, its search box is not where typing goes
Each service in its own window	No launcher needed, the service is already open	Set up once per service, and mail or chat has to be worth a window

The last row is the one people arrive at late. It is worth stating plainly because Chrome supports part of it directly: a site can be installed as an app from More, then Cast, save, and share, then Install page as app. Chrome documents that web apps installed this way can offer notifications, icon badges, offline storage and file system access. A service that lives in its own window never needs to be launched from a new tab, because it was never closed.

When the new tab page is the wrong thing to fix

There is a version of this problem that no new tab page solves, and it is common enough to check for before installing anything.

If the honest answer to "what gets opened after Cmd + T" is a fixed set of six or eight web applications, then the new tab page is being used as an application launcher, and the tiles are the least of the trouble. The real constraint sits underneath: a single browser profile holds one signed in session per service. Two accounts on the same chat tool, two mailboxes on the same provider, two dashboards for two clients, and one of them is always the one that is signed out.

Chrome's documented answer to that is a second profile. Profiles keep bookmarks, history, passwords and other settings separate, a new one is added from the profile icon at the top right with Add Chrome profile, and switching means picking a profile from the same menu. It works, and the cost is that switching is a whole window, so a quick look at the other account is never quick.

The category built around this constraint keeps each service in its own container instead, so the same web application can hold two signed in sessions side by side and neither one has to be closed. What that involves is set out on the Features page, grouping those containers by client or by role rather than by service is covered under Workspaces, and the list of services that load cleanly in that arrangement is on Supported apps.

That is a different problem from the one a new tab extension solves. Someone whose new tab is mostly a search box does not need any of it. Someone whose new tab is a launcher for eight logins has been solving the wrong half.

A quick way to tell which group applies: look at how many tabs are open right now and how many of them were opened this morning and never closed. A browser holding forty tabs is not a browser with a bad new tab page. It is a browser being used as an operating system, and the tiles on the first screen are a symptom of that rather than a cause. Fixing the first screen in that situation buys a week of feeling organised and then the count climbs back.

A two day test that settles it

Before installing anything, keep a rough count for two working days of what actually happens after a new tab opens. Three patterns cover almost everyone, and each one has a different answer.

If most new tabs end in typing a search or an address, leave the page alone. The address bar is already the fastest path, and any page drawn in the middle is one more thing to load and one more click to reach the box. Turning the cards off and picking My shortcuts is enough.

If most new tabs end at the same handful of sites, install those sites as apps or give them their own windows. The tiles were a workaround for services that should not have needed launching.

If new tabs are genuinely a moment of pause, a place to check a reading queue, a note or a calendar, then a dashboard extension earns its place. Choose it on permissions and on how fast it paints, not on the screenshot.

What to change first

Spend five minutes in Customize Chrome first: shortcuts set to My shortcuts, cards off, and a background that is not distracting. That removes the irritation without installing anything or granting any permissions.

Then look at what is left. If the remaining problem is a stack of web applications that keep signing each other out, that is a container problem rather than a new tab problem, and SpaceDeck is built around that unit.