What is different about marketing an online store
Most businesses advertise a handful of offers. An online store advertises its whole range, often thousands of items, each with its own price, stock level and margin. That changes where the marketing work happens. The product feed is the creative brief, the targeting signal and the bid unit at once, so the quality of titles, images, categories and identifiers decides which searches and feeds a product appears in. The shopping and catalogue ads guide makes the case that the feed is the real campaign; this guide adds that the feed is also where a store meets its first compliance test.
The second difference is competition on identical products. When several sellers list the same item, the shopper can compare them in one results page, so price, delivery promise, returns policy and reviews do much of the persuading. A store that cannot win on those has to win on range, content, service or brand, and its marketing has to say which. The marketplaces versus own store comparison sets out the trade between the reach a marketplace lends and the customer relationship only an own store keeps.
The third difference is when profit arrives. The first order often barely covers what it cost to acquire the buyer once product cost, delivery, payment fees and returns are counted. The store earns its margin when the customer comes back, which is why LTVTotal value a customer brings over time. against CACCost to acquire one new customer. is the comparison that decides how much a store can afford to pay for a new customer.
How e-commerce shoppers decide
Buying cycles in e-commerce range from minutes to weeks, and the range matters more than the average. A low-price replenishment item is bought on the first visit or not at all; a sofa, a bicycle or a laptop is researched across several sessions, devices and sites before anyone pays. Three patterns recur.
- Search-led purchases. The shopper knows what they want and searches for it by name, type or problem. Shopping ads, product listings in organic results and marketplace search do most of the work, and price, availability and delivery decide the click.
- Discovery-led purchases. The shopper was not looking until a social feed, a creator or a friend showed them the product. Catalogue ads, short video and creator content create the demand that search later captures, often days later and on another device.
- Considered purchases. The shopper compares specifications, reads reviews and returns several times. Detailed product pages, comparison content, remarketing within permission and a clear returns promise carry the decision.
Most stores sell in all three modes, product by product. Planning by mode rather than by channel avoids judging discovery activity on the same-session sales that only search-led products produce.
Constraints: regulation, platform policy and consent
The constraints below are the ones that most often shape an e-commerce plan. They differ by country and change over time; the linked sources are the authority, and this section explains how they affect marketing rather than what the law requires in a given case.
Product data has to match the store
Shopping platforms take product data from a feed and check it against the landing page. Google's Merchant Center product data specification defines the attributes a product needs, including the identifiers, price and availability the shopper sees, and products whose data disagrees with the page can be disapproved. In practice this makes the feed a controlled document: price changes, sale periods and stock-outs must reach the feed as quickly as they reach the page. The product feeds guide covers how to keep the two in step.
Prices, promotions and reviews are consumer protection matters
Claims about prices and discounts, how long an offer lasts, and what reviews say are regulated as consumer protection rather than as marketing style. In the UK the Competition and Markets Authority enforces consumer law and the Advertising Standards Authority administers the advertising codes; in the US the Federal Trade Commission's business guidance covers advertising claims and reviews. For a store this means a "was" price must be a price the item genuinely sold at, urgency messages must be true, and reviews may not be bought, filtered to hide negative ones or written by the business. The reviews and social proof guide explains what a store can and cannot do to collect them.
Creators and affiliates must disclose
Stores rely on creators and affiliate partners more than most industries, and both are endorsements. The FTC's Endorsement Guides expect a material connection, such as payment or free product, to be disclosed clearly, and the ASA applies the same principle to ads in the UK. The brand remains responsible for what partners publish on its behalf, so disclosure terms belong in every creator contract, as the contracts and disclosure guide sets out.
Email and text messages need permission
Email and text are the cheapest repeat-purchase channels a store has, and they are permission-based. The ICO's direct marketing guidance explains when electronic marketing needs consent and the limited case in which a business may market similar products to existing customers who were given a chance to opt out. In the US the CAN-SPAM compliance guide sets rules on identification, honest subject lines and honouring opt-outs. The e-commerce email and SMS flows guide builds the flows within those limits.
Tracking depends on consent
Remarketing audiences, conversion tracking and attribution all process personal data about shoppers. Under the GDPR that processing needs a lawful basis, and where consent is the basis, tags have to respect the shopper's choice. Google documents consent mode as the way its tags adjust to that choice. The consequence for a store is that some sales are never observed by the ad platforms, which report a mix of observed and modelled conversionsConversions a platform estimates statistically for activity it could not observe directly, such as unconsented or cross-device journeys.. The privacy, consent and measurement guide explains what is lost and what can be recovered.
Large catalogues need technical search hygiene
Organic search can be a store's largest free source of visits, but catalogues create duplicate and near-empty pages through filters, sorting and out-of-stock items. Google's e-commerce SEO guidance covers how stores should structure product and category pages, and the e-commerce SEO hub applies it.
Channel fit for e-commerce
| Channel | Role in e-commerce | Why |
|---|---|---|
| Paid search and shopping ads | Primary | Captures search-led demand with the product, price and image in the ad; the feed decides relevance |
| Paid social and catalogue ads | Primary | Creates discovery-led demand and retargets viewed products within consent |
| Email and SMS | Primary for repeat purchase | The cheapest route to the second order, limited to people who gave permission |
| Organic search | Primary over time | Category and product pages earn visits without a click price once the catalogue is crawlable |
| Marketplaces | Supporting or primary, by range | Reach among shoppers already buying, at the cost of fees and customer data |
| Creators and affiliates | Supporting | Borrowed trust and reach, paid on content or on sales, with disclosure duties |
| Video | Supporting | Demonstrates products that need showing; strongest for considered and discovery-led items |
| Content | Supporting | Buying guides and comparisons for considered purchases |
| Events | Rarely | Pop-ups and fairs suit few online-only stores beyond launches |
For tactics and budget posture by channel, use the online sales objective hub and its channel pages for paid search, paid social, email, marketplaces and affiliate. The platform setups are in the Google Ads sales playbook, the Meta Ads sales playbook and the TikTok sales playbook. For the content row, the guide to buying guides for online stores covers choosing, writing and linking guides so that they feed category pages rather than compete with them.
KPIs for an online store
| KPI | What it tells you | Role |
|---|---|---|
| Contribution after marketing cost | Whether sales leave money once product, fulfilment, returns and media are paid | Outcome |
| New customer acquisition cost | What a first-time buyer costs, separated from returning buyers | Outcome driver |
| Repeat purchase rate by cohort | Whether buyers acquired in a period come back | Outcome driver |
| CVRShare of visits or clicks that convert. by product type | Whether pages and offers convert the traffic they receive | Driver |
| AOVAverage revenue per order. | How much each order carries to cover fixed costs per order | Driver |
| ROASRevenue generated per unit of ad spend. against Break-even ROASMinimum ROAS to avoid a loss. | Whether a campaign clears the point where it stops losing money | Guardrail |
| Return and refund rate | Whether marketing is selling products people keep | Guardrail |
A KPI tree links these measures: contribution sits at the top, fed by new customers, repeat purchase and order value, each fed by the channel metrics beneath. The contribution margin and break-even guide shows how to derive the break-even return from a store's own margin.
Seasonality and the trading calendar
E-commerce demand follows a calendar of gifting periods, sale events, paydays and category seasons, and the calendar differs by country and range. Three planning habits follow. First, build stock, feed and creative changes before a peak, because shopping platforms and bidding systems need time to learn from new products and prices. Second, read results across the whole cycle, since conversion rates and costs move together during peaks and a return that looks poor in a pre-peak week may be paying for demand that converts later. Third, plan retention for the weeks after a peak, when a store holds its largest group of new customers and the second order decides whether peak spending was worth it. For search, the guide to seasonal and sale event pages keeps one permanent page per event so it carries its links from one peak to the next.
Data and measurement limits
Three limits apply to almost every store. Platforms count conversions their own way, with different windows and attribution models, so the totals across platforms exceed the orders the store actually took; the GA4 e-commerce tracking guide and server-side signals such as the Conversions API improve the observation but do not remove the overlap. Consent removes some shoppers from observation altogether, so modelled figures should be labelled as such. And returns arrive after the platforms have reported the sale, so return on ad spend in the ad account is always gross of refunds. The honest read is the store's own order data, net of returns, compared with periods or regions where spend changed.
A worked example
Quillmoor Paper Co., an invented online stationery store
Quillmoor Paper Co. sells notebooks, pens and desk accessories online, with an average order of £34 and a contribution margin of 45% after product, packaging and payment fees, so its break-even return on ad spend is about 2.2. Its shopping ads report a return of 3.0, but one in ten orders is returned, which lowers the net return to about 2.7. Repeat buyers place a second order within four months in roughly a third of cases. Quillmoor decides it can pay up to £18 to acquire a new customer, above the first-order contribution, because the second order covers the gap, and it moves its email sign-up offer from a discount on the first order to free delivery on the second, which protects margin and gives the repeat purchase a reason to happen.
Guides to use next
Start with the e-commerce marketing hub for the acquisition, conversion and retention levers, then the Google Shopping and Performance Max hub and the Meta commerce hub for the product advertising platforms. The product page guide and the repeat purchase and loyalty guide cover the two levers that most often decide whether acquisition was affordable, and the retention objective hub frames repeat purchase as a goal of its own. Other industries are in the industry library.
