What Is an Ecommerce Business?
An ecommerce business sells products or services online, taking payment through a website, app, or marketplace rather than across a counter. That definition is simple enough to be almost useless because it covers everything from a one-person Etsy shop to a wholesaler turning over eight figures through a trade portal.
What matters for growth is not the label but the mechanics: which model you operate, where your revenue actually comes from, and which of those sources you can move. This section covers all three before we get to any marketing channel, because choosing channels before understanding the mechanics is how most budgets get wasted.
How Ecommerce Actually Works
Every ecommerce transaction has the same four parts. Someone discovers your product, decides to buy it, pays for it, and receives it. Marketing works on the first two. Operations handle the last two. Most growth problems that look like marketing problems are actually failures in the second part, and we will come back to that.
Where businesses differ is the model, and the model changes what growth looks like.
| Model | How it works | Margin profile | Main growth lever | Typical constraint |
|---|---|---|---|---|
| Direct to consumer | You sell your own products through your own site | Highest, you keep the full margin | Traffic and retention | You have to generate all your own demand |
| Marketplace | You sell through Amazon, eBay, Etsy, or similar | Squeezed by fees and commission | Visibility within the platform | You do not own the customer relationship |
| B2B ecommerce | Trade customers order through a portal or account system | Lower per unit, far higher order values | Order frequency and account growth | Long sales cycles and approval processes |
| Social commerce | Selling within TikTok Shop, Instagram or similar | Fees plus content production costs | Content volume and creator reach | Platform dependence and algorithm shifts |
| Subscription | Customers pay repeatedly for recurring delivery | Strong lifetime value, high upfront acquisition cost | Reducing churn | One bad month of churn undoes a good month of acquisition |
Most established businesses run more than one of these at once, which is where the complications start. A brand selling direct and through Amazon is competing with itself on price, and the marketing that works for one actively undermines the other if nobody has decided which channel owns which customer.
What Actually Drives Ecommerce Revenue
Ecommerce revenue is not a mystery. It is arithmetic, and it has exactly four inputs:
Revenue = Traffic × Conversion rate × Average order value × Purchase frequency
Traffic is how many people reach your store. Conversion rate is the proportion who buy. Average order value is what they spend per order. Purchase frequency is how often they come back.
That is the whole thing. Every pound of ecommerce revenue anyone has ever made came from those four numbers multiplied together, and every growth tactic in existence works by moving at least one of them.
The reason this matters is that the four are not equally easy or equally expensive to move, and almost nobody treats them equally. Ask a store owner how they plan to grow, and the answer is nearly always about traffic. More ads, more SEO, more social. Traffic is the most visible lever, the most heavily marketed to business owners, and the most expensive one.
Meanwhile, conversion rate, order value, and frequency sit untouched. A store converting at 1.2% that gets to 1.8% has increased revenue by 50% with no additional traffic and no additional acquisition cost. A store that raises average order value by £6 through better bundling has done the same. These are not marginal optimisations; they are the same size of win as a major traffic increase, available at a fraction of the cost.
The practical use of the equation is diagnostic. Before you spend anything, work out which of your four numbers is weakest relative to your market. That is your constraint, and it is where money should go first.
What Growth Actually Requires
There are only three ways an ecommerce business grows: more customers, more spent per order, or more orders per customer. Every strategy reduces to some combination of those.
What growth requires beyond that is a set of conditions most stores have not met.
It requires unit economics that work. If it costs you £34 to acquire a customer who delivers £28 of gross margin over their lifetime with you, growth makes your situation worse rather than better. Scaling a business that loses money per customer simply loses money faster. This is the single most common reason apparently successful ecommerce brands collapse, and no amount of marketing skill fixes it.
It requires a store that converts. Sending more traffic to a page that does not persuade anyone is paying for the privilege of being rejected at volume.
It requires one identified constraint rather than four simultaneous projects. Stores that try to improve everything at once improve nothing measurably, because they cannot tell which change did what.
And it requires patience matched to the channel. Some channels produce sales this week. Others take two quarters before they produce anything at all. Mixing those timelines without planning for them is how marketing budgets get cut three months before they would have worked.
Why Most Ecommerce Stores Never Grow
You will see figures claiming that 80% or 90% of ecommerce businesses fail within their first few months. Those numbers circulate widely and have no credible source behind them, so we are not going to repeat them. What is well established is more useful anyway: most stores that survive do not grow. They plateau, often within the first year, and stay there.
The reasons are consistent, and they are rarely about the product.
- No identified constraint. The store works on whichever lever feels most urgent this month rather than the one that is actually limiting revenue
- Traffic without conversion. Acquisition spend rises, conversion rate stays flat, and cost per sale climbs until the channel is switched off and blamed
- Dependence on one channel. One algorithm change, one rising cost per click, or one account suspension removes most of the revenue overnight
- Discounting into margin collapse. Sales targets get hit through promotions, customers learn to wait for the next one, and full-price selling never returns
- No measurement worth the name. Platform-reported figures are taken at face value, channels are credited with sales they did not cause, and nobody can say what is working
The pattern underneath all five is the same. Each is a decision made without knowing which number it was supposed to move. That is why the equation above comes before any discussion of channels, and why the rest of this guide keeps referring back to it.
The Role of Digital Marketing in Ecommerce Growth
Digital marketing is how an ecommerce business moves the four numbers deliberately rather than hoping they improve. Each channel below does something specific: some bring people who are already looking, some create demand among people who were not, and some increase what existing customers are worth. None of them do all three, which is why channel choice is a strategic decision rather than a shopping list.
The ten channels that follow are the ones that reliably matter for UK ecommerce. For each, the first paragraph explains its impact and the second explains how it grows the business. Read the ones relevant to your constraint rather than all ten.
SEO for Ecommerce
Organic search is where most product research begins, and its impact is that the traffic arrives pre-qualified. Someone searching for a specific product, size, or specification has already decided they want the category and is now choosing between sellers. Unlike paid channels, the cost per visit falls over time rather than rising, because rankings earned once continue to deliver without per-click payment.
For growth, SEO works primarily on traffic, but the better return is often on conversion rate. Ranking well for precise, high-intent queries brings visitors who are closer to buying, so the same conversion work produces more sales. Category pages, product page structure, and the technical health of the site all feed this. The trade-off is time: ecommerce SEO is a two-to-three-quarter investment before it carries meaningful revenue, which makes it the wrong first move for a store that needs sales this month and the right one for a store that intends to exist in two years.
Google Search Ads
Paid search puts you in front of people at the exact moment they are looking for what you sell. Its impact is immediacy and control: you can be visible for a commercial query within hours, test which products and messages actually convert, and switch spend off the moment a campaign stops working. No other channel gives you that combination of intent and speed.
Its growth contribution is almost entirely traffic, and traffic of the highest available quality. That makes it the fastest way to establish whether your store converts at all, which is information worth having before committing to slower channels. What it will not do is get cheaper. Cost per click in competitive UK retail categories rises steadily, so search ads reward stores with healthy margins and punish those competing on price alone. Used well, it is a reliable revenue tap. Used as the only channel, it is a cost base that grows with your revenue rather than falling behind it.
Google Shopping and Performance Max
Shopping ads change the nature of the click. Instead of text, the searcher sees your product, your price, and your image before deciding, which means the people who click have already accepted your price. The impact is a higher-quality click and, usually, a markedly better conversion rate than text search for the same product. Performance Max extends the same product feed across Search, Shopping, YouTube, Gmail, and the Display network from a single campaign.
For growth, this moves traffic and conversion rate together, which is rare. The critical dependency is your product feed: titles, images, attributes, and stock accuracy determine what you show for and how often. Feed quality is genuinely the highest-leverage technical work in ecommerce paid media, and it is routinely neglected. Performance Max adds reach but removes visibility over where spend goes, so it suits stores with clean data and enough volume to let it learn, and frustrates smaller catalogues that need to control placement.
Paid Social Advertising
Meta, TikTok and Pinterest ads reach people who were not looking for you at all. That is their impact and their difficulty. You are interrupting rather than answering, so the creative carries the entire burden, and the same budget can produce wildly different results depending on whether the ad stops the scroll. Paid social is the main route to demand for products people do not know to search for.
It grows the business by expanding the top of the funnel, which moves traffic, and by enabling offers that lift average order value such as bundles and thresholds. Its real strength is creating demand that later shows up as branded search and direct visits, which is also why it is so often under-credited in reporting. The constraint is creative supply: performance decays as audiences see the same assets, so paid social is a production commitment as much as a media budget. Stores that cannot sustain new creative should not build their growth on it.
Organic Social Media
Organic social rarely sells directly, and judging it on last-click sales is why so many brands abandon it. Its impact is on the decision rather than the transaction. It is where a prospective customer checks whether you are real, whether other people buy from you, and whether the product looks as good outside the studio as it does on the product page.
The growth contribution is to conversion rate and purchase frequency rather than traffic. A healthy social presence reduces the hesitation that loses sales at checkout, and it keeps existing customers in contact with the brand between purchases, which is what makes them buy again. It also feeds branded search, the cheapest and highest-converting traffic any store has. The honest limitation is that it compounds slowly and demands consistency, so it belongs in the plan of a brand building something durable, not in the plan of a store that needs a result this quarter.
YouTube Advertising
YouTube is where consideration happens for anything that needs explaining, demonstrating, or justifying. Its impact is on products where a static image cannot carry the case: technical items, higher-priced purchases, anything where the buyer needs to understand how it works before they will risk the money. Video does in thirty seconds what a product page struggles to do in six hundred words.
For growth, it mainly creates demand, showing up later as search and direct traffic rather than as immediate purchases. It also lifts average order value for considered purchases, because a buyer who understands the product is more willing to choose the better specification. It is genuinely poor as a last-click channel and will look like a failure in any report that measures it that way. YouTube suits stores with something to demonstrate and a margin that tolerates a longer payback, and it wastes money on impulse products that never needed explaining.
Email Marketing
Email is the only channel in this list you own outright. No platform can raise its price, change its algorithm, or suspend your account. Its impact is the highest return per pound of any ecommerce channel, and the reason is simple: you are contacting people who have already bought from you or already told you they are interested.
It grows the business through the two levers almost everyone ignores: purchase frequency and average order value. Automated flows do most of the work: welcome sequences, abandoned basket and browse recovery, post-purchase follow-up, replenishment reminders, win-back campaigns. These run once built and produce revenue with no incremental acquisition cost, which means email improves your margin rather than just your turnover. The limitation is that it cannot create an audience, only monetise one, so it multiplies the value of every other channel rather than replacing any of them.
WhatsApp Marketing
WhatsApp gets read. Open rates far exceed email, because the message lands in the same place as messages from friends and family, and that is exactly why it has to be used carefully. Its impact is immediacy for things that genuinely need it, and irritation for anything that does not. The same intimacy that makes it effective makes it unforgiving of overuse.
For growth, it works on conversion rate and purchase frequency. Abandoned basket recovery, order and delivery updates, back-in-stock alerts and reorder prompts all perform strongly, and two-way conversation resolves the pre-purchase questions that would otherwise end in a closed tab. It also carries real compliance weight in the UK: consent must be explicit and opt-out must be easy, and getting that wrong is a regulatory problem rather than a marketing one. Used as a service channel that happens to sell, it is excellent. Used as a broadcast channel, it costs you the list.
Marketplace Advertising
Advertising on Amazon, eBay, or Etsy puts you in front of shoppers whose intent is not just commercial but transactional. They are not researching; they are choosing. The impact is a conversion rate no other paid channel reaches, because the entire environment exists for buying and the payment friction has already been removed.
It grows revenue through traffic and conversion simultaneously, and it grows it quickly. The costs are strategic rather than financial. You are renting visibility in someone else’s store; you do not own the customer data; fees compress your margin; and you are training customers to shop for your category on a platform where competitors appear beside you. For most brands, the right answer is presence on marketplaces for reach with a deliberate effort to move repeat buyers onto owned channels, rather than either ignoring marketplaces or depending on them.
Influencer and Creator Marketing
Creator content borrows trust you have not built yet. Its impact is on belief rather than awareness: a recommendation from someone a buyer already follows carries weight no brand advertising achieves, particularly in categories where people want to see the product used by someone who is not being paid to smile at it.
For growth, it creates demand and supports conversion, and the assets usually outlive the campaign, since the best-performing creator content becomes your paid social creative afterwards. That reuse is often worth more than the original post. The risks are real: rates rarely correlate with results, follower counts say nothing about purchase influence, and UK disclosure rules require clear labelling of paid partnerships. Treated as a content and credibility investment measured over months, it works. Treated as a one-off purchase of a post, it is the least predictable line in the budget.
What Digital Marketing Actually Delivers
The question every store owner actually wants answered is when. Here is a realistic view, and it is deliberately less flattering than most agency timelines.
| Channel | First signs | Meaningful results | What it takes to get there |
|---|---|---|---|
| SEO | 2 to 3 months | 6 to 12 months | Technical fixes, category and product page work, content, links. Continuous |
| Google Search Ads | Days | 4 to 6 weeks once stabilised | Budget, tight campaign structure, ongoing search term management |
| Shopping and Performance Max | 1 to 2 weeks | 6 to 8 weeks after the learning period | A clean, complete, accurate product feed above all else |
| Paid social | Days for data | 4 to 8 weeks to reach profitability | Continuous creative production, not just media spend |
| Organic social | 3 to 6 months | 6 to 12 months and beyond | Consistent posting and a reason for people to follow |
| YouTube ads | 4 to 8 weeks | 3 to 6 months | Video production plus tolerance for delayed attribution |
| First send | 1 to 3 months once flows are live | Automation setup, then light ongoing work | |
| Days | 4 to 8 weeks | Opt-in list building and disciplined message frequency | |
| Marketplace ads | Days | 4 to 8 weeks | Competitive pricing, strong listings, review volume |
| Influencer and creator | Per campaign | 3 to 6 months of repeated activity | Partner selection, briefing, and reuse of the content |
Two things about this table matter more than the rows themselves.
The first is that nothing here delivers meaningful results in under a month except paid channels, and paid channels deliver them by spending money rather than by building anything. A store with no budget and no patience has no viable strategy, and it is better to know that at the start than four months in.
The second is the mistake almost every store makes when reading a table like this: assuming the fast channels are therefore the best ones. They are not better; they are just faster, and they stop the moment you stop paying. The slow channels are the ones that lower your cost of sale over time. A serious plan runs both, funding the slow channels out of the returns from the fast ones, and judges each on its own timeline rather than holding SEO to a paid search reporting cycle.
The Ecommerce Growth Equation at a Glance
Here is the argument for working on all four levers rather than only the loudest one.
Improve traffic by 10%, and you have 10% more revenue. Improve all four levers by 10% each, and you do not have 40% more revenue. You have roughly 46%, because the four numbers multiply rather than add. Push each lever by 20% and revenue roughly doubles.
Now the reverse. Triple your traffic while conversion rate, average order value, and purchase frequency stay exactly where they are, and you have tripled your revenue and tripled your acquisition cost. Your margin per sale has not moved, your cost of sale has not fallen, and you are three times more exposed to the channel that delivered the traffic.
That is the whole case for a balanced approach, and it explains why the cheapest growth available to most stores is not a new channel at all. It is fixing the conversion rate, order value, and repeat purchase rate of the traffic they already pay for.
Building a System Rather Than Running Campaigns
There is a difference between a store that runs marketing and a store that has a marketing system, and it shows up around month four.
The store running campaigns launches things. A paid social push, then a sale, then some SEO work, then a creator collaboration when the paid social plateaus. Each is reasonable in isolation. Together they produce a pattern where nothing runs long enough to be judged, results cannot be attributed to any decision, and the team is permanently busy launching rather than improving.
The store with a system has decided what happens on what rhythm, who owns each thing, and what evidence would change the plan. That sounds bureaucratic, and it is the opposite: it is what frees you from reinventing the month every month.
A workable operating rhythm looks like this:
- Weekly: check spend against return by channel, pause anything clearly failing, review search terms and creative performance, note what to test next
- Every two weeks: ship new creative for paid social and review what the previous batch taught you
- Monthly: report against the constraint you identified, not against every metric available. Review conversion rate, order value and repeat rate alongside traffic
- Monthly: audit the product feed for accuracy, stock status, and missing attributes, since this quietly decays and takes paid performance with it
- Quarterly: review the channel mix itself. What should get more, what should get less, what should stop
- Quarterly: revisit unit economics. Acquisition cost, contribution margin, and payback period, because these drift and nobody notices until cash does
Two rules make the system hold. One thing changes at a time on any given channel, or you cannot attribute the result. And every recurring item has a named owner, because shared ownership of a weekly task means it happens for three weeks and then stops.
Measuring What Your Marketing Is Actually Doing
Most ecommerce reporting is a collection of numbers that individually look fine and collectively add up to more revenue than the business actually made. That is not a metaphor. Add up what Meta, Google, and your email platform each claim to have generated, and the total routinely exceeds real turnover, because each one counts the same sale.
Here are the numbers worth tracking and, more importantly, where each one lies to you.
| Metric | What it tells you | How it is calculated | Where it misleads |
|---|---|---|---|
| Customer acquisition cost | What it costs to win one new customer | Total acquisition spend divided by new customers in the period | Often calculated on all customers rather than new ones, which flatters it badly |
| Customer lifetime value | What a customer is worth over the relationship | Contribution margin per order multiplied by orders per customer over their lifetime | Frequently calculated on revenue rather than margin, which can overstate it several times over |
| Platform-reported ROAS | What a platform believes it generated | Platform-attributed revenue divided by spend on that platform | Each platform claims credit for sales the others also claim. Never sum these figures |
| Blended ROAS | Total revenue against total marketing spend | Total revenue divided by total marketing spend | Hides which channel is working, so it is a health check rather than a decision tool |
| Marketing efficiency ratio | How hard your whole marketing effort works | Total revenue divided by total marketing spend, tracked as a trend | Only meaningful over time. A single month tells you almost nothing |
| Payback period | How long before an acquired customer has repaid what they cost | Months until cumulative contribution margin exceeds acquisition cost | The metric most often ignored, and the one that determines whether growth is fundable |
Payback period deserves particular attention, because it is where profitable-looking businesses run out of money. If a customer costs £40 to acquire and returns £10 of contribution margin per month, the payback is four months. Grow quickly, and you are funding four months of that gap for every new customer, all at once. Plenty of brands have scaled themselves into insolvency doing exactly that while every ROAS figure on the dashboard looked healthy.
The practical approach is simple. Use blended figures and marketing efficiency ratio to judge whether marketing overall is working. Use platform figures only to compare a channel against its own past performance, never against another channel. And check payback period before deciding to scale anything.
What AI Search Means for Online Stores
Product discovery is moving into AI surfaces, and the data behind that is no longer speculative. Adobe recorded a roughly 1,200% rise in traffic to US retail sites from generative AI sources between July 2024 and February 2025, and a further increase of nearly 700% year on year over the 2025 holiday season. By March 2026, Adobe’s data showed AI-referred traffic converting around 42% better than non-AI traffic, reversing a substantial deficit recorded twelve months earlier.
The important part of that is not the growth rate but the reversal. AI-referred visitors were initially worse customers than search visitors. Now they are better ones, because someone who arrives after asking an assistant for a comparison has already done their research and narrowed the field.
Most of this data is US-based, and UK-specific figures remain thin, so treat the numbers as directional rather than as a forecast for your own store. The behaviour is the same on both sides of the Atlantic, but the scale may not be.
What has genuinely changed is how a product gets chosen. Traditional search showed ten links and let the buyer decide. An assistant presents two or three options and explains why. Being one of those options is the new version of ranking on page one, with a considerably narrower shortlist, and the criteria are different. Assistants and shopping agents read structured product data, not persuasive copy. Google’s Shopping Graph holds tens of billions of product listings fed largely from Merchant Center, and a listing missing key attributes is not ranked lower; it is simply not considered.
The practical response is unglamorous, and it is mostly data work:
- Complete your product feed properly: GTINs, brand, accurate categories, full attributes, real-time price and stock. Incomplete feeds get skipped rather than demoted
- Implement product schema on every product page: price, availability, reviews, and specifications in machine-readable form
- Build genuine review depth: assistants weigh volume and substance of reviews heavily when comparing options
- Write content that answers rather than ranks: comparison, sizing, compatibility and use-case questions, since those are exactly the queries assistants synthesise from
The checkout side of this is moving fast and reporting on it conflicts. Google has agentic checkout running in AI Mode and Gemini with limited US retail partners, and the protocols behind in-chat purchasing have changed more than once during 2026. Rather than chase that, note the stable part: shoppers still largely prefer to complete the purchase on the retailer’s own site even when an assistant made the recommendation. The discovery and consideration stages are what have moved. Your job is to be recommendable, and that is a data and reputation problem rather than a checkout one.
Building Your Ecommerce Marketing Strategy From Here
Work out which of the four numbers is limiting your revenue. Choose the channels that move that number, on timelines you can actually fund. Build the operating rhythm so the work continues after the enthusiasm fades. Measure with blended figures and check payback before you scale. Get your product data clean, because it now determines whether you appear in the places buyers are starting to look.
That sequence is the strategy. Most of it costs attention rather than budget.
If you would rather have it built and run by a team that does this daily, Innov8th works with ecommerce businesses across SEO, paid media, social, and conversion. Get in touch, and we will start with your numbers rather than our channel list.
Frequently Asked Questions
What is ecommerce in simple words?
Ecommerce is buying and selling online. A business lists products or services on a website, app, or marketplace, takes payment digitally, and delivers to the customer. It covers everything from a single-person craft shop to wholesalers taking six-figure trade orders through an account portal.
Is ecommerce still growing?
Yes, though not at pandemic-era rates. Growth has settled into a steadier pattern, with the share of retail happening online continuing to rise gradually. The bigger change now is where discovery happens, as AI assistants and marketplaces take a growing share of product research.
How many ecommerce businesses fail?
Widely quoted figures of 80% or 90% have no credible source and should be treated with suspicion. What is better established is that most surviving stores plateau rather than fail outright, usually because they never identify which number is limiting their revenue.
How do I increase sales in my ecommerce business?
Work out which of traffic, conversion rate, average order value, or purchase frequency is weakest, then fix that one. Most stores default to buying more traffic, which is the most expensive option. Conversion and repeat purchase work is usually cheaper and faster.
Which digital marketing channel works best for ecommerce?
There is no universal answer, because channels move different levers. Paid search and marketplace ads capture existing demand fastest. Paid social creates new demand. Email and WhatsApp increase what existing customers are worth. Your constraint decides which one matters.
How much should an ecommerce business spend on marketing?
Spend is better derived from your economics than from a percentage benchmark. Work out your contribution margin per order, your acceptable acquisition cost, and how long you can fund the payback period. That produces a defensible budget. A flat percentage of revenue does not.
How long does ecommerce SEO take to work?
Expect early signals in two to three months and meaningful revenue contribution in six to twelve. Technical fixes can show faster. Anyone promising competitive rankings in weeks is either working in a very soft niche or not being straight with you.
Is ecommerce profitable in the UK?
It can be, and margin discipline decides it rather than turnover. UK operators face competitive paid media costs, high delivery and returns expectations, and VAT considerations. Profitable stores tend to be those with clear positioning and repeat purchases, not the lowest price.
What is a good ROAS for ecommerce?
There is no single figure, because it depends entirely on your gross margin. Your break-even ROAS is roughly one divided by your gross margin, so a 40% margin needs about 2.5 to break even. A 4x ROAS is excellent for one store and loss-making for another.
Do I need an agency to grow an ecommerce store?
Not necessarily. Plenty of stores grow well in-house, particularly on one or two channels. An agency earns its fee when you need several channels running properly at once, when you lack time rather than ideas, or when you need someone to tell you which number to fix first.