Chromeのプロファイル: the setup order that holds up

Creating a Chrome profile takes about thirty seconds and no instructions are really needed for it. What takes longer is undoing a set of profiles that was created without a plan, which is the state most Macs arrive at after a year: five entries in the switcher, two of them named after email addresses nobody can read at that size, one that exists because a client asked for a shared login once, and no memory of which extensions are installed where. The steps below are ordered by how hard each decision is to reverse later, hardest first, because that is the only ordering that matters here.

Settle two questions before the second profile exists

The first profile on a Mac is not really a decision. It is the one Chrome created on first launch and it is stored in a folder named Default. The second profile is where a system starts, and two things should be decided before it is created rather than after.

The first is what a profile represents. Two answers are available and mixing them is what produces the unreadable switcher. A profile can represent a person or account, which is the intuitive choice: one for work, one for personal, one per client. Or it can represent a context, meaning the kind of work being done rather than whose credentials are in use. The account model is easier to start and grows one profile per new account, which is why the list reaches seven. The context model stays small and requires deciding in advance which accounts belong together.

The second is whether profiles are a boundary or a convenience. If the goal is keeping two sessions of the same service alive at once, a profile does that completely. If the goal is keeping anything genuinely away from someone else using the same Mac, a profile does not do that at all, since switching requires no password. Deciding which of the two is wanted prevents building an elaborate arrangement on a guarantee that was never offered.

Both questions are cheap now and expensive in six months, because by then each profile holds sessions, extension configuration and site permissions that do not move between profiles.

The creation steps, and the three fields inside them

Google sets out the sequence plainly in its help documentation:

On your computer, open Chrome. At the top right, select Profile Add Chrome profile. If you choose to sign in to Chrome with your Google Account, your profile name will automatically be your Account name. Choose a name, photo, and color scheme. Source: support.google.com

Three fields appear during that flow and each one is worth more attention than it usually gets.

The name is what shows in the switcher list and in the window title. It can be changed at any time and changing it costs nothing.

The photo is a small circle at the top right corner of the window. At that size a photograph is unreadable, so the useful choices are the plain avatars or nothing at all.

The colour scheme is the one field that does real work. It tints the browser frame, which means the entire top edge of the window carries the profile identity rather than a twenty pixel circle. Choosing colours that are hard to confuse with each other, rather than colours that look pleasant together, is what makes a five profile setup navigable at a glance.

If the same screen is used to sign into a Google Account, the name field is overwritten by the account name, which is the first thing that goes wrong for most people and is covered below.

A naming rule that still works at the fifth profile

Names are set during creation and almost never revisited, so they end up describing the moment the profile was created rather than what it does now.

Three rules keep the list readable. Use the role, not the address: Work, Client A, Reading. An email address in the switcher is unreadable at that size and stops being informative the moment two accounts share a domain. Keep names short enough to read in the window title, which means one or two words. Do not number them, since Profile 2 tells nobody what is inside.

There is a detail behind the names that causes confusion when looking at the file system. The visible name and the folder name are unrelated. Renaming a profile writes the new label into a small record in a file called Local State, while the folder on disk keeps whatever name it was born with. A profile displayed as Work can live in a folder called Profile 7. That mismatch is normal and not a fault, but it matters when copying profile folders between machines, because the folder names carry no meaning at all.

To rename an existing profile, open the profile menu at the top right, choose to customise the current profile, and edit the name there. Changes save immediately and nothing inside the profile is affected.

Sign in or stay local, decided at creation

The sign in prompt appears during creation, which is the worst moment to think about it, so the answer should already exist.

Signing a Google Account into a profile does two things. It turns on sync for bookmarks, passwords, history and settings, so they reach other devices signed into the same account. It also replaces the profile name with the account name, which undoes whatever naming rule was just applied and has to be set back manually afterwards.

Staying signed out keeps the profile fully functional. Cookies, extensions, bookmarks and passwords all work locally. Nothing leaves the machine and nothing arrives from another device.

The decision follows from what the profile is for. A profile that exists to hold a long lived work identity across two machines wants sync. A profile that exists to keep one client's console separate, or to test something, does not, and signing it in adds a copy of that data to an account for no benefit.

One warning applies to managed accounts. If the account belongs to a Google Workspace domain, signing it into a profile can bring administrator policy with it, which may control extensions, sync categories and other settings in that profile. That is the intended behaviour of managed profiles and it is reversible only by removing the account.

Deciding what is on screen at launch

A profile opens with either a blank new tab or a set of pages, and setting that per profile is what makes each one feel like a different workspace rather than the same browser with different cookies.

Startup pages are configured inside each profile, in the section of settings covering behaviour on startup. Three or four addresses is the useful range: the tools that are genuinely opened every morning. Ten is a slow launch and a wall of tabs that gets closed before it is read.

Pinned tabs are the companion setting. A pinned tab shrinks to its icon, sits at the left of the tab strip and returns when the profile reopens. Pinning the two or three applications that stay open all day, and leaving everything else unpinned, gives the tab strip a stable left edge that survives the daily churn.

There is one browser level option worth knowing, described in the same help documentation. Under the profile menu, the manage profile screen has a checkbox to show the profile chooser on startup. With it on, Chrome asks which profile to open rather than reopening the last one. That is helpful on a machine where the wrong profile keeps opening by accident, and an extra click every morning otherwise.

Extensions, downloads and search, set once per profile

Everything in this group is configured separately in every profile, which is a feature when it is deliberate and a chore when it is forgotten.

Extensions install per profile. A password manager added in the work profile does not appear in the personal one and has to be installed again. That is worth using rather than fighting: give each profile only the extensions its work requires. A profile that exists for reading does not need meeting tools, and each extension omitted is one less process running and one less set of permissions granted over browsing data.

The download location is also per profile, and it is the highest value item in this section because it is the one that fixes a real daily confusion. In each profile, open the downloads section of settings and point it at a folder belonging to that profile. Files then arrive already separated, without any sorting afterwards.

The default search engine is per profile as well. This is the only practical way to have a work profile search a company wiki by default while a personal profile searches the web normally.

Giving each profile a fixed place to open from

Once profiles are configured, the remaining friction is the trip through the menu at the top right, and macOS can remove most of it.

Chrome accepts a command line argument naming which profile folder to open. Running open -na "Google Chrome" --args --profile-directory="Profile 1" opens that specific profile directly, and the same command can be wrapped in a small application through the Script Editor or Automator, given an icon, and dropped in the Dock. Each profile then has its own icon in a fixed position.

Finding the right folder name takes one step, since the folder names carry no meaning. Open a profile, type chrome://version in the address bar, and read the profile path at the bottom of the page. The last component of that path is the value to pass.

Two limits are worth stating. These launchers open a profile, they do not move an already open window to the front in a predictable way, so the benefit is largest in the morning and smaller later in the day. And the Dock still groups every Chrome window under one application icon regardless of how the windows were opened, which is a constraint of how macOS treats the browser rather than anything configuration can change. Tools built as separate windows per application avoid that grouping entirely, as described on the workspaces page.

What the setup does not fix

A well ordered set of profiles removes the ambiguity about where things live. It does not reduce how often the crossing happens. Every answer that needs the other account is still a window change, and the count of those changes is set by the work rather than by the configuration.

That is the honest limit of this exercise. Naming, colours and launchers make a five profile machine legible. They do not make it fewer than five windows, and if the daily total of switches is in the dozens, the constraint has moved from organisation to architecture, which is the question the features page covers.

What to change first

Set a distinct frame colour on every profile and give each one its own download folder, since those two take five minutes total and remove the two most frequent confusions. Then count profile switches for a couple of days, and if the number stays high after everything is labelled correctly, look at whether a browser that keeps each web app in its own window, such as SpaceDeck, suits the way the day actually runs.

Frequently asked questions

Can a Chrome profile be renamed after it is created?

Yes, and nothing inside it is affected. Open the profile menu at the top right, choose to customise the profile, and edit the name. The folder on disk keeps its original name, which is why a profile called Work can sit in a folder called Profile 7.

Why does switching profiles open another window instead of changing the current one?

Because each profile is a separate browser session with its own processes, so Chrome opens a new window for it and leaves the previous one where it was. Closing the old window after switching is the only way to keep the count down.

Do extensions have to be installed again in every profile?

Yes. Extensions, their settings and their permissions all live inside a single profile. Signing a profile into a Google Account with sync enabled will reinstall the extension list on other devices using that account, but it does not share extensions with other profiles on the same Mac.

Is it better to show the profile chooser at startup?

It depends on how often the wrong profile opens by accident. The checkbox in the manage profile screen makes Chrome ask which profile to open instead of reopening the last one, which helps on a shared or heavily divided machine and costs an extra click every morning otherwise.

Should a Google Account be signed into every profile?

Only where sync is actually wanted. Signing in enables syncing of bookmarks, passwords and settings and renames the profile to the account name, and if the account is managed by a Workspace administrator it can also apply policy to that profile. A profile that stays local keeps full functionality without any of that.

Back to all posts