If you run an ecommerce business, you’ve probably heard someone say “just do SEO like a regular website.” They were wrong. That advice costs money.
Standard business websites, a law firm, a plumbing contractor, a SaaS company, have a handful of important pages. Maybe 20. Maybe 100 if they’re thorough. Search engines crawl them once a month and move on. Simple.
Your ecommerce site has thousands of product pages. Some are near-duplicates. Filters create 50 different URLs showing the same five products sorted by price. You’ve got regional variants, seasonal stock, out-of-stock items that still need visibility. Search engines can waste their entire crawl budget on your site without indexing anything useful. And if they index the wrong version of your product page, the filtered one, not the canonical, your traffic disappears.
That’s not a SEO problem in the traditional sense. It’s an architecture problem. And it requires a fundamentally different approach.
The Four Problems Standard SEO Doesn’t Solve
1. Crawl budget waste
Google allocates a crawl budget per domain. If your faceted navigation, sorting parameters, and session IDs create 100,000 variations of essentially the same page, Google will waste days crawling duplicates instead of finding new products or content that actually drives sales.
What a standard business website never faces: They have no filters. No product variations. No “sort by price ascending/descending” creating four separate URLs.
2. Duplicate product pages at scale
You have one blue shirt in three sizes and four colours. That’s 12 product URLs. If you own a chain of shops, you might have the same shirt across seven locations. Now it’s 84 URLs.
Do they all deserve separate rankings? No. Should Google index all 84? No. But which one should be canonical? Which should Google show in search results? And what happens to the customer who finds the wrong variant, out of stock at their local store, in stock 50 miles away?
What this breaks: Standard canonical tags work for one canonical URL per page. Ecommerce needs to handle product groups where multiple valid URLs exist for legitimate business reasons.
3. Faceted navigation and filters
Filters are brilliant for customers. They’re a SEO nightmare.
A customer clicks “blue,” then “size 10,” then “under £20.” They get a unique URL. Google sees a new page. Is it duplicate content or a unique product selection worth indexing? A standard business website doesn’t have this problem because it has no filters.
4. Structured data complexity
A product page needs structured data: price, availability, ratings, SKU, shipping info. But you have variants. You have multiple currencies (if you sell cross-border). You have multiple sellers or locations. Standard schema markup wasn’t designed for this scale.
Without clean structured data, you won’t appear in Google Shopping, AI-driven product comparisons, or emerging agentic commerce (where AI agents browse your site on behalf of customers). That’s not a ranking penalty, it’s invisibility.
What You’re Actually Choosing: Three Approaches
1. Do It Yourself (In-House Team)
The pitch: You know your business. You know your platform. Why pay an agency?
What you get:
- Full control of timelines and implementation
- Easier communication with your dev team
- Ability to iterate quickly when something breaks
What it costs:
- You need an experienced technical SEO person (or two). In the UK, that’s £45k–£70k+ per year for someone competent, more if they know ecommerce specifically.
- They need time to learn your platform’s architecture, your product data structure, your hosting setup.
- If they leave, knowledge walks out the door.
- You’re building best practices from scratch instead of using battle-tested templates.
The honest reality: Most UK ecommerce teams choose this because they think it’s cheaper. It’s rarely cheaper. It’s often slower. And one person can’t stay current with Google’s updates and manage crawl budgets and handle faceted navigation and fix structured data across 50,000 products. That’s three jobs minimum.
Best suited for: Teams with 2+ dedicated people, a straightforward technical setup, and time to learn. Not right for anyone cash-strapped or short-staffed.
2. Freelance or Small Agency
The pitch: Cheaper than a full agency, still gets done faster than in-house.
What you get:
- Fixed monthly cost (usually £1,500–£5,000 depending on scope)
- Someone who’s seen 20+ ecommerce sites and knows common pitfalls
- Access to templates and tools you’d have to build yourself
What it costs:
- You get 10–20 hours per month, not full-time availability
- Most are generalists, not ecommerce-specific
- Accountability is softer. Bad month? They miss targets with less consequence than an employee or agency.
- If your site has a weird technical issue, they may not have seen it before.
The honest reality: This works well for sites with straightforward technical needs, basic canonicalisation, improving crawl efficiency, fixing missing product schema. It breaks down when you hit the messy stuff: multi-variant products, regional/currency handling, or migrating to a new platform.
Best suited for: Smaller stores (under £500k revenue) with standard platform setups and clear, bounded problems.
3. Digital Agency (Full-Service)
The pitch: We fix everything. We own the results.
What you get:
- A dedicated account manager
- Access to people with 5+ years of ecommerce-specific technical SEO
- Audit, strategy, prioritised roadmap, ongoing monitoring
- Shared accountability for results
What it costs:
- Monthly retainer: £3,000–£10,000+ depending on site complexity and region
- Slower communication (you’re one client among many)
- Less flexibility. You can’t just add an hour here and there; you’re buying a service package
- Higher overhead means higher prices
The honest reality: Good agencies are expensive because good technical SEO is hard. Bad ones are cheaper and deliver bad resultso you pay less but get nothing.
A decent UK agency won’t promise “top 10 rankings in three months” for ecommerce technical SEO, because technical SEO doesn’t work that way. It fixes crawlability, structure, and data quality, which enables ranking gains. But it’s not a ranking tactic in itself.
Best suited for: Sites doing £2m+ revenue, complex setups (multiple currencies, marketplaces, high product volume), or where you’ve already spent money on technical problems and got nowhere.
The Real Comparison: What Gets Fixed
To be useful, here’s what each approach actually delivers:
| Problem | In-House | Freelance | Agency |
| Canonicalisation at scale | ✓ (if they know it) | ✓ Partial | ✓ Thorough |
| Crawl budget optimisation | ✓ (steep learning curve) | ✗ Often misses this | ✓ Standard practice |
| Product structured data | ✓ (if your platform helps) | ✓ If straightforward | ✓ + validation |
| Faceted nav strategy | △ (technical, not strategic) | △ Band-aid fixes | ✓ Architecture-level |
| Ongoing monitoring | △ (if resourced) | ✗ Usually not | ✓ Included |
| Cross-functional buy-in | ✗ One person can’t push this | ✗ No authority | ✓ From day one |
A Quick Reality Check: What Each Actually Solves
Technical SEO for ecommerce will NOT:
- Rank you #1 for “buy cheap trainers online” (that’s keyword strategy + content + links)
- Fix bad product photography or missing reviews (that’s UX and product)
- Solve inventory problems (that’s operations)
It WILL:
- Stop Google wasting crawl time on filter parameters you don’t need indexed
- Ensure your best product variant ranks, not a filtered dead-end
- Get your products into Google Shopping and comparison engines (which drive 30–40% of ecommerce revenue)
- Reduce the chance of silent indexation failures that tank traffic without warning
Which Should You Actually Choose?
Ask yourself three questions:
1. How much revenue are we losing to poor discoverability right now?
- If you don’t know, get an audit first. A 2-day ecommerce SEO audit costs £1,500–£2,500 and tells you exactly what’s broken.
- If you’re losing £5k+ per month, an agency pays for itself in year one.
- If you’re losing £500/month, a freelancer might be enough.
2. Do we have time to learn this ourselves?
- Ecommerce technical SEO has a steep learning curve. If your team is already stretched, you won’t finish it. Honest answer: most UK SME retailers are stretched.
3. How complex is our setup?
- Single-region, single currency, simple platform? Freelance is fine.
- Multi-region, variants, inventory syncing, cross-selling logic? You need someone who’s seen this before. That’s an agency.
What to Do Next
If you think you have a technical SEO problem:
1. Get a free crawl audit using Screaming Frog or similar. Look for:
- How many pages does Google think you have?
- How many are actually indexed?
- Are there obvious duplicates?
2. Check Google Search Console for crawl errors, crawl stats, and coverage issues. If you see red flags there, you have a real problem.
3. Depending on what you find, either:
- Brief a freelancer for a one-off fix (£2k–£4k)
- Request an audit and strategy from an agency (£2k–£3k, no obligation to hire them after)
- Invest in hiring someone in-house if you’re large enough (£50k+, multi-year commitment)
The worst choice is doing nothing. Silent technical issues in ecommerce don’t improve. They compound. Every month your faceted navigation wastes crawl budget is another month your new products don’t get indexed.