Pick the smaller tool

Browser Extension vs Web App: A Fair Decision Table, Including the Times a Bookmarklet Will Do

Somebody in the meeting says the product needs an extension. Heads nod. Nobody asks what it would do that a web page cannot, which is the only question that matters, and the one that saves the most money when the answer is nothing.

November 17, 2026 · 9 min read

acmefreight.com · build decision worksheet
Fit scores for one job: quoting loads from a carrier portal
The job lives on someone else's page, so the extension wins
Sample
Browser extension on the carrier portalStrong fit
Web app with copy and pasteWorks, slower
Bookmarklet for the dispatch teamToo rough
Existing store extensionNone found
Illustrative: a sample decision worksheet for a fictional freight broker

The short answer

Build a browser extension when the job has to happen on someone else's website, inside the browser, while the user is working there: showing data on a page, capturing it in one click, acting on the current tab. Build a web app when the job lives on your own site, must work on any device, or has to ship without store review. Many products need both, with the app doing the heavy lifting.

On this page
  1. The question that settles most cases: whose page is the user on?
  2. Extension, web app, bookmarklet: the side-by-side
  3. Jobs where an extension earns its place in the toolbar
  4. Where a web app beats any extension
  5. The cheap answers: a bookmarklet, a plain page or an existing extension
  6. Why GatorCrawl Keywords is a web app and an extension at once
  7. If you need both, build the app side first

The question that settles most cases: whose page is the user on?

A web app lives at your address. People come to it, sign in and work there. An extension lives in the browser and goes where the user goes. Chrome's getting started guide puts the extension's powers plainly: "customizing the user interface, observing browser events, and modifying the web." The last phrase is the one a web app cannot match. A web page runs on its own site; reading or changing someone else's page inside the user's browser is the job Chrome's content scripts exist to do.

So start with the user's day. If the moment you want to improve happens on Google, LinkedIn, a supplier's portal or a customer's site, an extension is how you put something there. If the moment happens anywhere, on a laptop at the office or a phone in a truck, a web app reaches it and an extension does not. Both are built from the same web technologies, HTML, CSS and JavaScript, per Chrome's docs, so the choice is about where the work happens, not which language it uses.

Extension, web app, bookmarklet: the side-by-side

Facts in the table come from the browser makers' own documentation, read in November 2026. The bookmarklet column is our practice: a bookmark that runs a few lines of JavaScript on the current page when clicked, with no store, no install and no permissions screen.

Browser extension, web app and bookmarklet compared, with store facts as of November 2026
Browser extensionWeb appBookmarklet
Works on other sites' pagesYes, through content scripts on the sites it declaresNo; it runs on your own siteOnly the page open when it is clicked
DevicesDesktop browsers; Safari web extensions also run on iOS 15 and later, inside an App Store appAny device with a browserDesktop browsers where bookmarks can hold scripts
Store review before releaseYes; Google says most reviews take a few days and some take up to a few weeksNone; you deploy when readyNone
Shipping a fixEvery update is reviewed againLive minutes after deployUsers replace the bookmark
Code loaded from your serverNot allowed; all logic ships in the packageNormalPossible, though we keep them self-contained
Install frictionStore listing plus a permission promptA link and a sign-inDrag a link to the bookmarks bar
Store feesChrome $5 once; Microsoft Edge none; Apple's developer program 99 USD a yearNoneNone
Runs in the backgroundBriefly; Chrome normally stops an idle service worker after 30 secondsOn your servers, as long as neededNo

Sources for the extension column: Chrome's review process page, the Manifest V3 overview on remote code, the service worker lifecycle page, the $5 fee as mentioned on Chrome's What's new page, Microsoft's Edge developer account page, and Apple's developer program page and Safari extensions documentation. The cross-browser guide covers what changes between stores.

Jobs where an extension earns its place in the toolbar

Data on top of a page you do not own

Keyword numbers on Google, prices beside a supplier's catalog, account history beside a support inbox. Our SEO extensions page shows the pattern.

One-click capture

Save the contact, the listing or the order you are looking at into your own system, without copy and paste. See CRM extensions.

Right-click lookups

Chrome's context menu API adds items for selected text, links or images, so a lookup is one click from wherever the user is reading.

A side panel that stays open

Chrome's side panel, available from Chrome 114 per its reference as of November 2026, can stay open across tabs if set to, which suits an AI assistant or a reference tool.

A tool every employee gets

Google's admin documentation describes force-installing extensions for an organization's users. See internal team extensions.

The common thread: the user is already on a page, and the extension saves a trip to another tab. If removing that trip does not change anyone's day, the extension is decoration.

Warm shallow-focus desk with an open laptop showing a blank glowing screen, a notebook of sketched boxes, a coffee cup and a hand holding a pencil
Most of this decision happens on paper, before anyone writes code.

Where a web app beats any extension

Plenty of ideas that arrive as extension requests are web apps in disguise. These are the signs.

  • The users are on phones. Field crews, drivers and customers mostly are. An extension will not follow them there, apart from Safari on iPhone through an app.
  • The job is a dashboard or a report. Tables, charts and exports want a full screen at a stable address, not a toolbar popup.
  • The work runs long. Chrome normally stops an idle extension service worker after 30 seconds and ends any single request that runs past 5 minutes, per its docs as of November 2026. Overnight imports belong on a server.
  • Many roles sign in. Admins, managers, staff and customers each with their own view is a web app's home ground.
  • Customers will not install anything. A link always beats an install prompt for a first visit.
  • You need to ship fixes on your schedule. A web app deploys when you press the button; an extension update waits for review.

If those describe your project, our web app development and custom software pages are the better starting point, and an extension can come later as a front door to the app.

The cheap answers: a bookmarklet, a plain page or an existing extension

Before anyone pays for a custom build, three cheaper answers deserve a hearing. A bookmarklet handles a one-off job on the current page, such as copying a few fields or opening the same record in your system, for a small team that will tolerate something rough. A plain web page with a form or a search box handles any job that does not need to sit on someone else's site. And an extension already in the store may do the job outright: our review of SEO Chrome extensions is a good model for judging one before you install it.

Why GatorCrawl Keywords is a web app and an extension at once

Our own product shows how the two divide the work. The GatorCrawl Keywords web app does the heavy part: accounts, plans, credits and the keyword data itself. The extension does one thing the app cannot. On a Google results page, its content script reads the query, the related searches and the People also ask questions, sends them to the service worker in one batch, and draws a compact bar under the search box with searches, CPC, KD and a 12-month trend line.

The extension holds no keyword data of its own. Its service worker carries the user's extension token and calls the web app's API once per results page. Its options page connects the account by opening the app's connect page, then shows the credits left. That split kept the extension to two permissions and one site, and it means pricing changes happen in the app without a store review. The case study walks through it part by part.

The extension is a window. The app is the house.

If you need both, build the app side first

When the answer is both, order matters. The extension depends on the app's accounts and API; the app does not depend on the extension. Building the API first also means the extension stays small, which keeps its permission list short and gives a reviewer less to read.

  1. 1

    Define the one job on the other site

    Write down the page, the moment and the click. If you cannot, revisit the web app column.

  2. 2

    Expose what the extension needs from the app

    Sign-in, the data, the balance. Connecting to an app or API you already run goes live the same day through our API integration service.

  3. 3

    Build the extension as a thin client

    Logic it must run ships in the package; prices, keys and data stay on the server.

  4. 4

    Submit under your own account

    Store submission, privacy disclosures and the permission plan are part of the build. Our review guide covers what to expect.

Extensions start from $1,500, depending on complexity, with a fixed written estimate before work starts; the build service and pricing guide explain what moves the number. If you are not sure which to build, a fractional CTO conversation is cheaper than either. The section overview lists everything we make. DataCram has built for the web since 1999 (about); call (210) 346-0848.

Questions owners ask about custom browser extensions

It depends on the job, not the format. A focused extension that calls an API you already have can be a modest build; DataCram's extensions start from $1,500, depending on complexity, as of November 2026. An extension that needs its own accounts, billing and data behind it is really an extension plus a web app, and costs accordingly.

Not inside the user's browser. A web app runs on its own site. Showing or changing content on someone else's page while the user browses it is what Chrome's documentation assigns to content scripts, which only extensions have. A web app can link to that site or pull data from its API, but it cannot draw on the page.

Apple documents one route. It says Safari web extensions are available on iOS 15 and later, built as an app extension and distributed through the App Store. We read no source on extensions in mobile Chrome, so we make no claim about it. If your users work on phones, a web app is the safer choice.

Not directly, but you can add one. The usual approach keeps the web app as it is and builds a small extension that signs in to the same account and calls the app's API. That is how GatorCrawl Keywords works. The extension does the one job that needs to happen on another site, and everything else stays in the app.

When a small team needs a one-off action on the current page, such as copying a few fields or opening a matching record in your system, and nobody minds something plain. There is no store, no install and no permission screen. When the job needs to run automatically, sign in, or reach customers, it is time for an extension.

Not always. A simple extension that changes how a page looks can work alone. Anything with accounts, payments, AI or private data should have a server, because Chrome requires all extension logic to ship inside the package and secrets do not belong there. Often the server already exists as your web app's API.

Keep reading

Free · No obligation

Get a straight answer on extension or app

Describe the job, where your users are when they need it, and what systems it touches. We reply with what we would build, or tell you a simpler option will do.

Which of these sounds like you? (pick any)

Rather talk now? Call (210) 346-0848.

CallGet a quote