Web Development The page the ad lands on can change this week.

Ecommerce, short for electronic commerce, is buying and selling over the internet. An ecommerce business is any business that takes orders and payment online, through its own website, a marketplace like Amazon or eBay, a social platform's shop, or all three at once. E-commerce, eCommerce, ecommerce: one word, three spellings, and "ecom" is what people in it actually say.

How it works, mechanically, because the mechanics are where the money disappears: a storefront shows products from a catalog; a shopper adds things to a cart; checkout collects shipping and payment; a payment gateway (Stripe, PayPal, Shopify Payments, Adyen) authorizes the card and moves the money, minus about 3%; the order goes to fulfillment, which picks, packs and ships it from your own stock, a third-party warehouse, or a marketplace's network; and the platform tracks the order, handles returns and emails the customer at every step. The shopper sees the storefront. The business lives in everything after the click.

The industry in 2026 has three kinds of player. Marketplaces aggregate thousands of sellers under one checkout; Amazon alone is roughly 40% of US online retail, which is a statement about leverage, not just size. Platforms (Shopify, WooCommerce, BigCommerce) give a brand its own store without building software. And the brands and retailers themselves sell through some mix of the two, usually arguing internally about which one is the "real" business. Online sales are about 16% to 17% of total US retail, climbing a point or so most years, far higher in electronics, books and apparel, lower in grocery and auto.

The models are named by who sells to whom: B2C, a business to consumers, most online retail; B2B, business to business, with quotes, accounts and bulk pricing; C2C, consumers to each other, eBay and Facebook Marketplace; D2C, direct-to-consumer, a brand selling on its own site instead of through retailers. Dropshipping is not a business model. It's a fulfillment arrangement where a supplier ships what you sold, which removes inventory risk and most of the margin with it.

Ecommerce is a checkout plus everything that happens after it. The storefront is the cheap part.

To start an online store, find a product you can source at a margin that survives shipping, payment fees and advertising, pick a platform, build the store, connect payments and shipping, and launch with a plan for how the first hundred customers will find it. The build is a few days. The business is the product and the traffic, and I've watched people spend six months on the build to avoid confronting the other two.

In order:

1. Product and economics first. Know your landed cost (product plus freight plus duty), your price, payment fees (about 3%), shipping to the customer, and what it will cost in advertising to acquire a customer. If the margin after all of that is thin on paper, it will be negative in practice, because the paper didn't include returns.
2. Platform. For most product businesses: Shopify. Hosted, reliable, payments and shipping built in, apps for everything else, from $39 a month. WooCommerce if you're already on WordPress and want control and lower fixed costs, at the price of running hosting, updates and security yourself. BigCommerce competes with Shopify on large catalogs. Wix and Squarespace for stores that are mostly a website with a few products. "Free" stores exist, Square Online, Ecwid's free tier, Big Cartel, with limits on products and features; a free store is for testing an idea, not for running a business.
3. Build. Domain, theme, product listings with real photos and complete descriptions, collections, shipping rates, tax, a payment provider, the legal pages (privacy, returns, terms), and a test order placed and refunded by you before anyone else places one.
4. Measurement, before launch. GA4 with purchase tracking, Google Search Console, and the pixel and conversion API for any ad platform you intend to use, installed before the first sale so the first sale is recorded. Most stores install this in month three and lose the data that would have told them what worked.
5. Launch and acquire. A store with no traffic sells nothing, and it will not sell anything "organically" for months. The first customers come from somewhere you chose on purpose: paid search on the product's own terms, paid social, an audience or list you already have, a marketplace alongside the site, or press. Organic search is built in parallel and arrives in year one if you're disciplined and lucky.

An ecommerce platform is the software running the store: catalog, cart, checkout, orders. An ecommerce website is the store as the customer sees it. The best platform is the one whose constraints you can live with: Shopify for ease and uptime, WooCommerce for control, BigCommerce for big catalogs without transaction fees, and a custom build only if the business has genuinely strange requirements and an engineering team to keep it alive.

Getting into ecommerce as a career instead of as a founder means learning one platform deeply, one acquisition channel deeply, and the analytics that connect them. The field pays people who can make a store sell. It does not pay people who can set one up; a sixteen-year-old can set one up.

Economics, then platform, then measurement, then traffic. The build is the easiest part and the one everyone starts with.

Conversion rate optimization is raising the share of visitors who do the thing the site exists for, buy, sign up, request a quote, by changing the site rather than by buying more traffic. CRO stands for conversion rate optimization; "conversion optimization" is the same thing. The conversion rate is conversions divided by visitors: 20,000 visitors and 400 orders is 2%, which is about the median for US ecommerce. Good stores run 3% to 4%. The best run higher and don't tell you how.

CRO is the highest-leverage lever most businesses own and the one they spend the least on, because it isn't a channel and no vendor sends a rep. It multiplies everything at once: a store that moves from 2% to 2.5% just grew revenue by a quarter without a dollar more on ads, and every future ad dollar now buys a quarter more. Nothing else in marketing has that shape. Which is why it's worth doing properly instead of by whoever has the strongest opinion in the room.

Properly means a testing program. Analytics and session recordings show where people leave: product page, cart, shipping step, payment form. Research, watching recordings, reading support tickets, surveying the people who didn't buy, produces hypotheses about why. Hypotheses become A/B tests: half the visitors see the change, half don't, run until the sample is big enough to trust, which on most sites means weeks per test and a few hundred conversions per variation. Most tests lose or say nothing. The program wins because the few that win are kept and stacked. The ways it fails are always the same: calling tests early because someone wanted a result for the Monday meeting, testing on too little traffic to read, and changing four things at once so nothing can be credited.

What actually moves conversion on an ecommerce site is dull and I'll list it so you can check your own: speed, product photography and real detail, shipping cost and delivery date visible before checkout, guest checkout, fewer form fields, accelerated payment (Shop Pay, Apple Pay, PayPal), trust signals (reviews, a returns policy, a phone number), and removing everything that isn't the one thing the page is for.

Personalization is showing different people different things based on what you know: returning customers see their category, visitors from a specific ad see the matching offer, shoppers in a region see their shipping terms. It lifts conversion when the differences are real and the data behind them is reliable, and it's sold as far more than that by people who sell personalization software. The honest version is tested like anything else, and most of it loses to a faster page. Cart, checkout, product pages, forms, mobile, and the full method are in the CRO half of the web service.

A half-point of conversion is a quarter more revenue from the same traffic. It is the cheapest growth you will ever buy, and it's bought with tests, not opinions.

WordPress is free, open-source software for building and running websites, a content management system that runs more sites than anything else on earth: about 43% of the web in 2026, per W3Techs, from personal blogs to news publishers to enterprise sites to stores. "WP" and "word press" mean WordPress. It started in 2003 as blogging software and is now a general-purpose site builder that happens to still be excellent at blogging.

There are two things called WordPress, and the confusion between them generates most of the questions. WordPress.org is the software: free to download, free to use, open-source license; you install it on hosting you pay for and you own everything. WordPress.com is a commercial hosting service from Automattic that runs the same software for you, with a free plan on a wordpress.com subdomain and paid plans from a few dollars a month for a custom domain, plugins and commerce. When someone says they built a site "on WordPress," they almost always mean the .org software on their own hosting, and when they complain WordPress won't let them install a plugin, they're on .com's free plan.

How it works: the software is PHP with a MySQL database; a theme controls how the site looks; plugins add functionality, from a contact form to a full store; the block editor is where pages and posts get written. Since version 5.9 in 2022, block themes and the Site Editor let you edit the whole site, header and footer included, visually. WordPress is a CMS in the full sense: content separated from presentation, with an interface non-developers can run. Is it a CMS? Yes; it's the CMS.

What it costs: the software is $0. Hosting for a .org site runs $3 to $15 a month on shared hosting, $20 to $100 a month on managed WordPress hosting (WP Engine, Kinsta, Flywheel) that handles updates, caching, backups and security for you, and more for real traffic. A domain is $10 to $20 a year. Premium themes and plugins are optional and typically $30 to $300 a year each, and the total creeps up as plugins accumulate. WordPress.com's paid plans bundle hosting and run roughly $4 to $45 a month, with commerce higher.

WordPress is used for essentially everything: blogs, business sites, portfolios, news, membership and course sites, and stores through WooCommerce. WordPress.org does not host websites; you need a host. WordPress.com does. "Is WordPress down" usually means WordPress.com, whose status lives at wordpress.com/status; a self-hosted site being down is a hosting problem wearing a WordPress costume.

Is WordPress secure? The core software is, maintained by a dedicated security team and patched fast. The risk is the ecosystem: outdated plugins and themes, weak passwords, abandoned plugins with known holes, and the many, many sites nobody has updated since launch. A WordPress site kept updated, running a small number of maintained plugins, with strong passwords and two-factor auth on decent hosting, is as secure as anything. One that isn't is the most-attacked target on the web precisely because there are forty million of them and the attackers know most are unattended.

WordPress is free software that runs 43% of the web. The software is secure. The site you haven't updated since 2022 is not.

To build a website with WordPress: get hosting and a domain, install WordPress, pick a theme, make your pages, add the few plugins you actually need, publish. On any modern host the mechanics take an afternoon. The design and the writing take as long as you let them, which for most people is far too long before anything ships.

The steps:

1. Hosting and domain. Pick a host with one-click WordPress install, which is all of them now. Register a domain, through the host or elsewhere. Managed WordPress hosting costs more and does the maintenance; for a first site, shared hosting is fine and you upgrade when traffic says so.
2. Install. The host's installer asks for a site name, an admin username and a password and does the rest. The manual "famous five-minute install," uploading files and creating a database yourself, still exists and almost nobody needs it.
3. Theme. Appearance → Themes → Add New. The default theme that ships with WordPress is a fine starting point; the directory has thousands of free ones; premium themes are sold by their developers. A block theme lets you edit everything in the Site Editor without touching code.
4. Basics. Settings → General for the title and tagline. Settings → Permalinks, set to "Post name," so URLs are words and not ?p=42; do this before you publish anything, because changing it later breaks every link you've shared. Settings → Reading to choose a static home page or the blog.
5. Pages. Pages → Add New. Home, about, contact, and a services or products page is the minimum. The block editor adds headings, text, images, columns and buttons by inserting blocks.
6. Plugins, sparingly. Plugins → Add New. A form plugin, an SEO plugin, a backup plugin, caching if the host doesn't do it, a security plugin. That covers most sites. Every plugin is someone else's code running on your server; fewer is better and the site is faster.
7. Publish, then set up Search Console and Analytics; see SEO on WordPress.

Starting a blog on WordPress is the same build with posts instead of pages at the center: posts are dated, categorized, tagged entries shown newest-first on the blog page; pages are static. Set the blog as the home page or give it its own page in Settings → Reading, make categories that match what you'll actually write about, and write. The blog that fails is the one with twelve categories and three posts.

Developing a WordPress site, as opposed to building one, means custom work: a theme written for the site (a child theme of an existing one, or a block theme from scratch), custom plugins for things nothing off the shelf does, custom post types for structured content, integrations with other systems. That's developer work, and it's what separates a site that fits the business from a site that fits the template. How to use WordPress once it's live is the next section.

Hosting, install, theme, permalinks to "Post name," pages, five plugins, publish. Set permalinks before anything else or you'll regret it in a year.

Managing a WordPress site happens in the admin dashboard at yoursite.com/wp-admin or yoursite.com/wp-login.php, with the username and password set at install. Can't log in? "Lost your password?" emails a reset; if the email never comes, the host's control panel can reset it, and the database can be edited directly as the last resort. Some sites move the login URL with a security plugin, in which case whoever set it up knows where it went, and if they've left the company that's your first task.

Themes. Appearance → Themes: Activate switches, Add New installs. Changing a theme changes the look and keeps the content. Updating happens from Dashboard → Updates, and if someone customized the theme files directly, the update overwrites their work, which is why customizations belong in a child theme and why so many sites are stuck on a theme version from 2021. To find what theme any WordPress site is using, view the page source and look for /wp-content/themes/<name>/ in the stylesheet links, or use a detector site. The same move on a Shopify site: view source, search Shopify.theme.

Cache. Clearing or purging the cache tells the caching layer to rebuild its stored copies so your changes show up. Where the button is depends on what's caching: a plugin (WP Rocket, W3 Total Cache, LiteSpeed Cache, WP Super Cache) has one in the admin bar or its settings; a managed host has one in its panel; a CDN like Cloudflare has a purge in its dashboard; and your own browser needs a hard refresh. When a change doesn't appear, clear them in that order, and know that "the site didn't update" is a cache problem nine times in ten.

Footer. Depends on the theme type. Block theme: Appearance → Editor, open the Footer template part, edit it. Classic theme: Appearance → Customize → Footer, or Appearance → Widgets for footer widget areas, or the theme's own options panel. The copyright line that won't change from any of those is hardcoded in footer.php, which a child theme overrides.

Backups. Files and database, on a schedule, somewhere other than the same server. Managed hosts do this; otherwise UpdraftPlus, Jetpack Backup or BlogVault to cloud storage. Then test a restore once, because a backup you've never restored is a hope. Undoing a change to a page is built in: the editor keeps revisions in the post's settings panel and any version can be restored. Undoing a broken site after a bad update is what the backup is for.

Plugins. The must-have list is five long: backup, security (Wordfence or Solid Security), caching if the host doesn't, SEO (Yoast, Rank Math or All in One SEO), and forms (WPForms, Gravity Forms or Fluent Forms). Past that, install a plugin for a problem you actually have, and delete the ones you stop using. The site with 47 plugins is slow, and it's slow because of the plugins.

Speed. In order of impact: better hosting, page caching, image compression and modern formats (ShortPixel or Imagify), fewer and lighter plugins, a lighter theme, a CDN, deferring scripts the page doesn't need yet. Measure with PageSpeed Insights before and after, and the number that matters is the real-user Core Web Vitals, not the lab score that the agency screenshots.

Users and roles. Users → All Users. Five built-in roles: Administrator (everything), Editor (all content), Author (own posts), Contributor (own drafts, can't publish), Subscriber (profile only). Modifying what a role can do takes a plugin like Members or User Role Editor. Give people the smallest role that does their job, keep Administrator to one or two humans, and remove the agency's admin account when the contract ends.

Editing. Pages and posts in the block editor. Adding HTML to a page: the Custom HTML block. Adding code to the site's head or footer for analytics or verification: a plugin like WPCode or the SEO plugin's header settings, never the theme files. Showing recent posts on a page: the Latest Posts block or the Query Loop block, filtered by category, tag or author.

Securing it. Keep core, themes and plugins updated; strong unique passwords and two-factor; limit login attempts; HTTPS; one security plugin; delete unused themes and plugins; keep backups. Seven steps, and they prevent nearly every WordPress compromise I've ever been called in to clean up.

Updates, a child theme, five plugins, a tested backup and two-factor. Most "WordPress problems" are a site nobody's maintaining.

SEO on WordPress is SEO, with the platform-specific plumbing handled by a plugin. Out of the box WordPress is reasonably search-friendly: clean HTML, readable URLs once permalinks are set to "Post name," an automatic XML sitemap at /wp-sitemap.xml, and fast enough on real hosting. What it lacks natively is an interface for title tags, meta descriptions and the finer controls, and that's what Yoast SEO, Rank Math or All in One SEO add. Pick one. Not two.

Where the SEO title is in WordPress: with a plugin installed, every page and post gets a box under the editor (or a sidebar panel) with fields for the SEO title, the meta description and a preview of the search result. The SEO title becomes the page's <title>; without a plugin, WordPress uses the post title plus the site name. That box is also where people look for "where to add keywords in WordPress," and the answer is: not in a keywords field. In the SEO title, the URL slug (editable at the top of the editor), the H1 (the post title), the opening paragraph, the subheadings, image alt text (set in the media library or the image block), and the meta description for the click. The "focus keyword" field in Yoast and Rank Math scores how well the page uses the phrase against the plugin's checklist. It is not sent to Google. It does not affect ranking. It's a spellchecker for SEO, and the green light means the plugin is happy, not that Google is.

Adding meta keywords in WordPress without a plugin is possible by editing the theme's header, and pointless, for the reason the meta tags section explains at length. Adding SEO keywords to a WordPress site without a plugin is the on-page list above, none of which needs a plugin; the plugin makes the title and description fields convenient and nothing more.

Adding Google Analytics to WordPress: Google's own Site Kit plugin connects Analytics, Search Console and Ads from one screen, or you put the Google tag or a Tag Manager container in the head with a header-and-footer plugin or the SEO plugin's settings. Editing theme files for this is fragile and vanishes on the next theme update, which is how analytics silently stops on a third of the sites I audit.

Improving WordPress SEO beyond that, or improving the "score" or the "ranking," is the same list that moves any site: faster hosting and caching, compressed images, a lighter theme, one intent per page, real titles, internal links between related posts, and content that's better than what currently ranks. The score the plugin shows is the plugin's checklist. Green from Yoast is not a signal Google receives. What WordPress SEO is, in the end, is regular SEO with the plumbing already done and a plugin that gamifies the easy part.

The plugin gives you the title and description fields. Everything that actually ranks the page is the same as on any other site, and the green light isn't Google.

WooCommerce is the free, open-source ecommerce plugin for WordPress, built by Automattic, that turns a WordPress site into a store with products, cart, checkout, orders, shipping and tax. By store count it's the most-used ecommerce platform in the world, because it rides WordPress's installed base. The plugin is free. The costs are hosting, the payment processor's fees, and paid extensions, which is where the ecosystem makes its money: subscriptions, bookings, advanced shipping and the like run $49 to $299 a year each. A real WooCommerce store costs hosting plus a few hundred dollars a year in extensions, against Shopify's fixed monthly plan, and that comparison is the whole decision.

Setting it up: install the plugin from Plugins → Add New, run the wizard (store address, industry, product types), add products (simple, variable, grouped or downloadable), set shipping zones and rates, set tax, choose payment methods. Day to day is the Orders screen, stock from the product list, reports from Analytics. Installing it is installing any plugin; the wizard does the rest, and the rest takes an afternoon.

WooPayments is WooCommerce's own payment gateway, built on Stripe and run by Woo, which takes cards, Apple Pay and Google Pay without a separate processor account, at fees in line with Stripe's (2.9% plus 30 cents per US transaction in 2026). WooPay is the express checkout that comes with it: a shopper who's saved details with WooPay checks out on any WooPay-enabled store in a couple of clicks, the way Shop Pay works for Shopify. Setting up WooCommerce payments means installing WooPayments, connecting a Woo account and verifying the business; other gateways (Stripe direct, PayPal, Square, Authorize.net) are separate plugins and fine.

Creating a coupon code: Marketing → Coupons → Add Coupon; code, discount type (percentage, fixed cart, fixed product), amount, expiry, restrictions (minimum spend, specific products, one per customer). Exporting products: Products → Export produces a CSV of every product and attribute; Import on the same screen takes a CSV back, which is how catalogs get bulk-edited or migrated, and how they get corrupted by a spreadsheet that reformatted the SKU column as a number.

WooCommerce is free and the store costs hosting plus extensions. WooPayments is Stripe with a Woo badge. Export to CSV before any bulk edit, and don't let Excel touch the SKUs.

Shopify is a hosted ecommerce platform: a subscription that gives a business an online store with catalog, cart, checkout, payments, shipping and order management built in and running on Shopify's servers. Shopify was founded in 2006 in Ottawa, Ontario, by Tobias Lütke, Daniel Weinand and Scott Lake, after Lütke built the software to sell snowboards online and noticed the software was the better business. It's headquartered in Ottawa, publicly traded, and has been remote-first since 2020. "Qué es Shopify" gets the same answer in Spanish.

How it works for sellers: sign up, pick a plan, choose a theme (Dawn is the free default; the Theme Store sells others), add products and collections, connect a domain, turn on Shopify Payments or another gateway, set shipping, open the store. Everything runs from the admin on the web or in the Shopify app. The App Store adds reviews, email, subscriptions, upsells and a few thousand other things. Shopify handles hosting, security, updates and uptime; you handle products, marketing and fulfillment. Plans in 2026 start at $39 a month for Basic, with Shopify at $105 and Advanced at $399, plus Shopify Plus for enterprise from around $2,300 a month. Every sale pays card processing (2.9% plus 30 cents on Basic, less on higher plans) and, if you use a third-party gateway instead of Shopify Payments, an extra transaction fee on top, which is Shopify's way of telling you which gateway to use.

What Shopify is for: any business selling physical or digital products online that wants to own the customer relationship instead of renting it from a marketplace: direct-to-consumer brands, retailers adding online, and increasingly large enterprises on Plus. A "Shopify store" is just a store built on it. Is Shopify good, is it worth it? For most product businesses, yes, without much argument: it's the most reliable, best-supported way to run a store without engineers, the ecosystem is the biggest, and the checkout converts. It's worth less to businesses that mostly sell on marketplaces, that need deep customization the platform fights, or that are too small for the plan, which is the usual reason people leave and the usual reason they come back.

Shopify POS is the point-of-sale app and hardware for selling in person from the same inventory and customer records as the online store. The Shop app is Shopify's consumer app, owned by Shopify, free, where shoppers track orders from any Shopify store, find brands and pay with Shop Pay. Shopify Collective lets Shopify retailers list and sell other Shopify brands' products with the supplier fulfilling and the two splitting revenue. The Partner program at partners.shopify.com is for agencies, developers and designers who build stores, themes and apps; joining is free and partners earn referral and app revenue. "What happened with Shopify today" is status.shopify.com for outages and the investor relations page for the stock. How many Shopify stores exist is an estimate: Shopify reports millions of merchants; third-party crawls put live stores around four to five million worldwide in 2026.

Shopify is the default because it works, and the fees are the price of it working. The extra charge for a third-party gateway isn't a fee; it's a hint.

To build a Shopify store, start a trial at shopify.com and work through the admin's setup checklist: products, theme, domain, payments, shipping, checkout, then remove the password page to go live. A simple store is a day's work. The time goes into product photos, descriptions and how the collections are organized, which is also where the sales come from, so that's time well spent.

Step by step:

1. Account. Sign up with an email, name the store (changeable), answer the onboarding questions. A few days free, then a nominal first month.
2. Products. Products → Add product: title, description, images, price, compare-at price if on sale, inventory, weight for shipping, variants (size, color), and the search engine listing fields at the bottom that most people skip. Adding a product in the Shopify app is the same form on your phone, and yes, the whole store can be built and run from a phone; the theme editor is just easier with a mouse.
3. Collections. Products → Collections. Manual ones are hand-picked; automated ones use rules (tag, price, vendor). Collections become the navigation, so build them around how customers shop, not how your warehouse is organized.
4. Theme. Online Store → Themes. Dawn is free and fast; paid themes in the Theme Store run $150 to $400. Customize opens the visual editor where sections get added and arranged per template. Assigning a template to a page: create the new template in the editor (a product template with a size guide, say), then on the page or product in the admin, choose it under Theme template.
5. Domain. Settings → Domains. Buy through Shopify or connect one you own.
6. Payments. Settings → Payments. Shopify Payments where available is simplest and dodges the extra transaction fee; PayPal and others alongside.
7. Shipping. Settings → Shipping and delivery: zones, rates (flat, weight-based, carrier-calculated), local pickup if that's real for you.
8. Checkout and legal. Settings → Checkout for customer accounts and fields collected; Settings → Policies for refund, privacy, terms and shipping, which checkout links to.
9. Launch. Place a test order yourself. Then Online Store → Preferences, remove the password.

Setting up Shopify, starting a Shopify store or business, opening, creating, launching, making one: all the list above. Starting a business with Shopify adds everything that comes before the store, which is under starting an online store and the starting a business guide. How to use Shopify once it's running is the admin's Orders, Products, Customers, Analytics and Marketing screens, which explain themselves to anyone who got through the setup.

Nine steps and a test order. Build collections the way customers shop, not the way the warehouse is shelved.

Getting sales on a Shopify store is a traffic problem first and a conversion problem second, in that order, and Shopify solves neither. It never claimed to. A new store has no visitors, and the store owners who make it are the ones who decided, before launch, exactly where the first hundred customers were coming from and went and got them.

Traffic, in the order a new store should try it: paid search on the product's own terms through Google Ads and Shopping, because the intent is explicit and the feedback is fast; paid social on Meta, where the pixel and Conversions API let the platform find buyers once it's seen a few dozen purchases; an email list, started on day one with a reason to join and sent through Klaviyo or Shopify Email; marketplaces like Amazon alongside the store for products with existing demand; and organic search and content, which takes months and gets started in parallel instead of waited for. Marketing a Shopify store and promoting it are the same thing: running those channels, consistently, with the numbers in front of you.

Conversion: the apps a new store actually needs are three. Reviews (Judge.me or Loox), email (Klaviyo), and something for post-purchase upsell or bundles once there's volume to upsell. Shop Pay on, because accelerated checkout converts better and that's not a theory. Everything else, the pop-ups, the countdown timers, the loyalty widget, gets added for a reason, not by default, because every app slows the page and a slow page costs more conversion than the widget adds. Making the store itself convert better is the CRO section.

Decide where the first hundred customers come from before you launch. Three apps, Shop Pay on, and nothing that slows the page.

Shopify is good for SEO in the way that matters: it doesn't stop a store from ranking, and its defaults do most of the technical work for you. Every product, collection, page and post has editable title and meta description fields (the "Search engine listing" section at the bottom of each form), the XML sitemap generates itself at /sitemap.xml, canonical tags are set, the hosting is fast, and there's a blog. Stores on Shopify rank in every category. The platform is not the reason your store doesn't.

What it constrains, so you know: URL structure is fixed (/products/, /collections/, /pages/, /blogs/) and can't be flattened; products reached through a collection get a nested URL, which Shopify handles with a canonical to the plain product URL; robots.txt was locked for years and is now editable through a robots.txt.liquid template; and heavy app use slows pages, which hits Core Web Vitals. None of that prevents ranking. All of it comes up in audits as if it did.

Adding keywords to a Shopify site is the same as anywhere: the product or page title (that's the H1), the search engine listing title, the URL handle (editable in the same section), the description, image alt text (set per image), and collection descriptions, which are empty on most stores and are the easiest win you'll find. Getting a Shopify store to rank for keywords beyond that is plain SEO: collections built around how people search instead of around internal categories, product descriptions that aren't the manufacturer's copy duplicated across a hundred other stores, a blog that answers the questions around the product, and links from other sites. Improving or increasing SEO on Shopify, "doing SEO for a Shopify website," Shopify SEO in general: all that list plus the SEO reference.

Shopify won't stop you ranking. Your empty collection descriptions and your manufacturer's copy will.

There's no place to "shop on Shopify." Shopify is the software behind individual brands' stores, so buying from a Shopify store means buying from that brand on its own website, like any online shop: add to cart, check out, pay. The store may look like any other; the giveaways are a /products/ URL and a Shop Pay button at checkout.

To browse many Shopify stores in one place, the Shop app (free, from Shopify) lets shoppers discover and buy from Shopify brands, track orders from any of them, and pay with Shop Pay, which saves payment and shipping details for one-tap checkout on any store that offers it. Shop Pay is free for shoppers and is a large part of why Shopify checkouts convert. Beyond the Shop app there's no Shopify marketplace; shopping "on Shopify" or "in Shopify" is shopping at a store that happens to run on it, and the store would rather you didn't know.

You don't buy from Shopify. You buy from a brand that pays Shopify. The Shop app is the closest thing to a storefront Shopify has.

To tell whether a website is legitimate, check the things a scam site can't easily fake: how long the domain has existed, whether the business exists and can be reached outside its own site, what people say about it where the site has no control, and how it wants to be paid. No single check settles it. Three or four together do.

The checks, roughly in order of how much they tell you:

• HTTPS, the padlock, means the connection is encrypted. It does not mean the site is honest; the scam sites have it too. Its absence disqualifies; its presence proves nothing.
• Domain age. A WHOIS lookup (who.is, or ICANN's lookup) shows the registration date. A "trusted since 2009" store on a domain registered three weeks ago is lying to you, and now you know how to catch it.
• A real business behind it. A physical address, a phone number someone answers, a company name findable in a business registry, a social presence with history, the same contact details on more than one page.
• Reviews the site doesn't control. Trustpilot, Google reviews, the Better Business Bureau, Reddit. Testimonials on the site itself are decoration. A store claiming to be busy with no reviews anywhere else is a warning.
• How it takes money. Legitimate stores take cards and PayPal, which give you chargeback rights. A site that only takes wire transfer, crypto, Zelle or gift cards is a scam, with exceptions so rare you should act as if there aren't any.
• Prices and theater. Luxury goods at 80% off, everything "almost sold out," a countdown that resets when you reload.
• The craft. Product photos stolen from other brands (reverse image search finds them), copy with the errors of a template, policies that don't match the business.
• Safe Browsing. Google's Transparency Report (transparencyreport.google.com/safe-browsing/search) tells you whether Google has flagged it; ScamAdviser and URLVoid roll up several checks.

To check that a product is authentic rather than a site, buy from the brand's own store or from retailers the brand lists as authorized; a marketplace seller with no history offering a current-season product at half price is selling counterfeits or nothing. "What website is this" is answered by reading the address bar slowly, because lookalike domains (amaz0n, paypa1, the brand name with an odd suffix) are the standard phishing move and they work on people in a hurry.

For site owners, the same list read backwards is how to look legitimate, which you must, because visitors run these checks without knowing they're doing it: visible contact details, a real address, policies, outside reviews, normal payment methods, a domain that's been around. That's trust in the conversion sense, covered under CRO. Whether WordPress specifically is secure is under what WordPress is.

Check the domain age, the outside reviews and the payment methods. A site that only takes gift cards has told you everything.

Google Sites is Google's free website builder, part of Google Workspace and available to anyone with a Google account at sites.google.com. It builds simple pages by dragging in text, images and embedded Docs, Sheets, Maps, Calendars and YouTube videos, publishes to a sites.google.com/view/yourname address or a custom domain, and needs no hosting and no technical knowledge. For an internal team site, a project or class page, an event page, a basic brochure, it's fine and it's free.

What it can't do is grow: no ecommerce, no real blog, little design control, minimal SEO beyond page titles and descriptions, no plugins or custom code past embeds, and no way to move the site anywhere else without rebuilding it. A business that expects its website to find customers builds on WordPress, Shopify, Squarespace or Wix, and the ones that start on Google Sites rebuild within a year.

"Google website" sometimes means something else: the simple sites Google used to generate from Google Business Profile listings. Google shut those down in March 2024 and the addresses now redirect to the profile. A business that relied on one needs a real site, and the profile's website field should point at it, because right now it points at nothing.

Google Sites is for pages, not businesses. The Business Profile websites are gone, and if yours was one, your website field is empty.