E-commerce growth gets treated as a traffic problem by default — get more visitors, sales will follow. In practice, most established stores have far more room to grow from the levers already inside their existing traffic than from acquiring more of it.
The 6 levers, ranked by typical speed to impact: E-commerce Growth
| Lever | Typical time to see movement | Common fix |
|---|---|---|
| Conversion rate | Weeks | Checkout friction, page speed, product page clarity |
| Return rate | Weeks-months | 3D/AR product visualisation, better sizing information |
| Average order value | Weeks | Bundling, thresholds for free shipping, smarter upsells |
| Repeat purchase rate | Months | Email/SMS lifecycle flows, loyalty incentives |
| Organic traffic | Months | SEO, content, technical fixes |
| Paid efficiency | Weeks | Campaign structure, creative testing, landing page match |
Why returns are the most overlooked lever
Every returned order costs twice — once on the original sale, again on shipping and restocking. Platform data shows interactive 3D and AR product content converting up to 94% higher than static photos alone, with roughly 40% fewer returns, because shoppers resolve fit and appearance uncertainty before buying rather than after. See our AR product visualisation service for how this applies to a specific catalogue.
Fixing conversion before scaling traffic
Sending more traffic to a leaking checkout just increases the size of the leak. A 20% conversion rate improvement is mathematically identical to a 20% traffic increase — except it's usually far cheaper to achieve, since it's fixing what you already have rather than paying for more of it.
Where paid and organic fit together
PPC gives fast, controllable volume while SEO investment compounds toward a lower long-term cost per acquisition. Most growing stores run both simultaneously rather than choosing one — see PPC advertising and SEO services for the full detail on each.
The bottom line
E-commerce growth isn't one lever — it's usually the smallest, cheapest fix that's been ignored because "get more traffic" sounds like the obvious answer. Audit conversion and returns before increasing ad spend; it's almost always the faster route to more revenue.
Diagnose before you spend
The six levers are only useful once you know which one you are actually short on. Most stores can work that out in an afternoon with data they already have.
| What you observe | What it usually means | Where to look first |
|---|---|---|
| Traffic is fine, sales are not | Conversion problem | Product pages, checkout, delivery clarity |
| Sales happen but margin is thin | Returns or discounting | Return reasons, promotional dependence |
| Customers buy once and vanish | Retention problem | Post-purchase email, product mix |
| Ads worked, then stopped | Audience saturation or creative fatigue | Frequency, new creative, new channels |
| Nobody arrives at all | Genuine traffic problem | Organic visibility, paid, marketplaces |
| Baskets are small | Average order value problem | Bundling, thresholds, cross-sell |
Only the fifth row is genuinely a traffic problem, yet "we need more traffic" is the diagnosis most stores reach for. It is worth checking which row you are actually in before committing a budget.
Reading your own return data
Returns are the lever most stores never examine, because return data usually lives in a system nobody reports from. It is worth the effort, because return reasons tell you precisely what your product pages are failing to communicate.
Segment returns by reason and by product. A concentration of "not as described" or "looked different" points at photography and copy. A concentration of "doesn't fit" points at sizing information, and often at a single product line rather than the whole catalogue. "Changed my mind" is usually a pricing or expectation problem rather than a product one.
Once you have that split, the fix is usually specific and cheap: better images for one category, a size guide for one product line, clearer delivery timing on one page. That is a materially better use of budget than a general "improve the site" project.
The conversion checks worth doing first
Before any redesign, work through these on a real phone, on mobile data, as a customer who has never seen your site.
- Is the delivery cost visible before checkout? Unexpected delivery cost is a leading cause of abandonment, and it is free to fix.
- Can you buy without creating an account? Forced registration costs sales.
- Does the product page answer the obvious objection? Sizing, materials, compatibility, how long delivery takes.
- How many taps from product to paid? Count them. Each step loses people.
- Does the page load acceptably on mobile data? Not on office wifi.
- Is the returns policy easy to find? A clear policy increases conversion; a hidden one reads as a warning.
These are unglamorous and they routinely outperform a redesign, because they address the specific reasons individual customers left rather than the general feeling that the site looks dated.
Sequencing: what to fix in what order
Doing these in the wrong order is how growth budgets get spent without moving revenue.
- Measurement first. If you cannot see where people drop out, everything after this is guesswork. Get funnel tracking and return reasons visible before anything else.
- Then conversion and returns. Both compound: they improve the value of every visitor you already have and every visitor you will pay for later.
- Then retention. Post-purchase email and a reason to buy again. Cheaper than acquisition and it raises what you can afford to pay for a customer.
- Then average order value. Bundling and thresholds, once the basics convert.
- Then traffic. Paid and organic last, when the machine they feed is working.
The temptation is always to start at step five, because more traffic feels like progress and is easy to buy. But traffic sent into a leaking funnel is the most expensive way to discover you had a conversion problem.
What to measure, and how often
Weekly, watch conversion rate, average order value and the cost of acquiring a customer. Monthly, look at return rate by reason and by product, repeat purchase rate, and contribution margin after all variable costs including returns processing. Quarterly, look at customer lifetime value against acquisition cost, since that ratio decides how much you can responsibly spend on growth.
Two cautions. Never judge a change on a period shorter than one full purchase-and-return cycle, or returns will land after you have already declared victory. And segment by product category rather than reading a blended average, because a blended number hides exactly the concentration you are looking for.
Frequently Asked Questions
Which lever should an established store fix first?
Whichever has the biggest gap versus your category benchmark — for most stores that's conversion rate or return rate, since both compound: a 1-point conversion improvement affects every future visitor, not just this month's traffic.
Is it better to spend a growth budget on traffic or conversion first?
Conversion first, in almost every case — improving what happens to existing traffic is typically cheaper and faster than acquiring new traffic, and it also makes every pound spent on future traffic more productive.
How much does AR product visualisation typically cost for an e-commerce catalogue?
A single-product 3D/AR viewer typically runs £2,500-£6,000; a multi-product catalogue runs £6,000-£20,000+ depending on how many models need building from scratch versus existing assets.
How do I tell whether I have a traffic problem or a conversion problem?
Look at what you actually observe. Traffic arriving but not converting is a conversion problem, so examine product pages, checkout and delivery clarity. Sales happening on thin margin points at returns or discount dependence. Customers buying once and disappearing is retention. Ads that worked and then stopped suggest audience saturation or creative fatigue. Only genuinely low arrivals is a traffic problem — yet "we need more traffic" is the diagnosis most stores reach for first, which is why growth budgets so often get spent without moving revenue.
Why are returns described as an overlooked growth lever?
Because return data usually sits in a system nobody reports from, and because return reasons tell you precisely what your product pages fail to communicate. Segment returns by reason and by product. A concentration of "not as described" points at photography and copy. "Doesn't fit" points at sizing information, often on one product line rather than the whole catalogue. "Changed my mind" is usually pricing or expectation. The resulting fixes are specific and cheap — better images for one category, a size guide for one line — and they protect margin you have already earned.
In what order should I fix things?
Measurement first, because without visibility of where people drop out everything after is guesswork. Then conversion and returns, since both compound by improving the value of every visitor you already have and every one you will pay for later. Then retention, which is cheaper than acquisition and raises what you can afford to spend on it. Then average order value through bundling and thresholds. Traffic last, when the machine it feeds actually works. Starting with traffic is tempting because it feels like progress, but it is the most expensive way to discover a conversion problem.
Learn more from Google's official guidelines.
