Technical SEO Audit for E-commerce (2026)

A technical SEO audit is the foundation of any successful e-commerce SEO strategy. For online stores in Bangladesh — whether you run a Daraz storefront, a Shopify shop, a WooCommerce catalog, or a custom Next.js commerce build — getting the technical fundamentals right can mean the difference between page 1 and page 10. This 2026 checklist distills five years of e-commerce SEO consulting in Bangladesh for brands like Walton and HATIL into a prioritized 60-point audit you can run quarterly.

Why Technical SEO Matters More for E-commerce Than Any Other Vertical

E-commerce sites face unique SEO challenges that B2B or media sites rarely encounter at the same scale:

Thousands of SKUs with unique product pages, each a potential ranking surface

Faceted navigation (size, color, brand, price) that creates millions of crawlable URL combinations

Session IDs and parameter URLs that pollute the index

Out-of-stock and seasonal products that decay rapidly

Structured data complexity (Product, Offer, Review, AggregateRating, BreadcrumbList, ItemList)

JavaScript-rendered product carousels and filters that don't always get indexed

Massive duplicate content risk from product variations, sorting options, and pagination

A single misconfigured facet rule can multiply your crawl budget waste by 100x. A missing canonical on product variants can cause keyword cannibalization across hundreds of pages. This checklist covers every major technical risk for an e-commerce site in 2026.

Section 1: Crawlability & Indexation (12 Checks)

1.1 Robots.txt Configuration

Verify `robots.txt` allows crawling of `/product/`, `/category/`, and `/collections/` paths

Block parameter-only URLs (`?`, `&sort=`, `?color=` etc.) explicitly

Disallow staging, internal search results, and admin paths

Confirm XML sitemap is referenced in `robots.txt`

Avoid blanket `Disallow: /` on JS or CSS that Googlebot needs to render pages

1.2 XML Sitemap Health

Sitemap contains only canonical, indexable 200-status URLs

No noindex pages, no redirect chains, no 404s in sitemap

Separate sitemaps for products, categories, blog posts, and images

Each `<lastmod>` reflects the actual content update timestamp

Submit product sitemaps to Google Search Console after major catalog changes

1.3 Index Bloat Audit

Pull a full list of indexed URLs from GSC → Pages → Indexed pages

Compare against sitemap count; if indexed count > sitemap count by 30%+, you have index bloat

Identify and either canonicalize, noindex, or 404 the bloat (filter URLs, search results, internal admin paths)

E-commerce sites with 10k+ indexed pages often have 30-50% bloat

1.4 Crawl Budget Efficiency

For sites with 10k+ product pages, check `crawl stats` in GSC

Server log file analysis: which bots hit the site, how often, and on which sections

Identify low-value URL patterns consuming crawl budget (search results, tag pages, session IDs)

Compare Googlebot crawl rate vs organic traffic — a healthy ratio is ~1:1 to 1:3

1.5 JavaScript Rendering

Use `site:yourstore.com` and `Fetch as Google` (URL Inspection → Live Test) to confirm product pages render fully

Check whether product titles, prices, and reviews are visible in the raw HTML response (not just after JS execution)

For React/Next.js stores: verify Server-Side Rendering is enabled for product detail pages

Use the Rich Results Test on product URLs to confirm structured data is server-rendered

1.6 Canonical Tag Audit

Every product page has a self-referencing canonical

Variant pages (size, color) canonicalize to the parent product

Pagination uses either `rel=next/prev` (deprecated but still respected) or a `view-all` canonical

Faceted navigation URLs use `rel=canonical` pointing to the unfiltered page

No canonical chains (canonical pointing to a page that itself canonicalizes elsewhere)

1.7 Pagination

Either implement `view-all` canonical to a single page, or use paginated canonicals

Ensure pagination pages have unique title tags and meta descriptions (`Page 2 of 5 - Blue Shirts`)

Block low-value pagination (page 50+) from indexing via `noindex, follow`

Internal link pagination with descriptive anchor text (`Blue Shirts - Page 2`), not just "Next"

1.8 Soft 404 Detection

Identify URLs returning 200 status but rendering empty / "no results" / error content

Convert these to true 404 or 410, or render unique content

GSC → Pages report flags these under "Soft 404"

1.9 URL Structure

Product URLs follow a flat hierarchy: `/category/product-name` not `/category/subcategory/subsubcategory/product-name`

URLs use hyphens, lowercase, no underscores

No session IDs, no parameter-based sorting, no relative paths

URL length < 100 characters where possible

1.10 Hreflang for International Stores

If selling in multiple countries or languages, every URL has self-referencing and reciprocal hreflang

x-default points to the global or language-selection page

No mismatched hreflang (page A links to page B, but page B doesn't link back to A)

Currency and pricing structured data reflects the target region

1.11 Noindex Audit

No accidental `meta name="robots" content="noindex"` tag on product or category pages

`X-Robots-Tag` HTTP header is not set to `noindex` on any product page

No `noindex` in paginated category pages

Use GSC URL Inspection to spot-check 20 random product URLs

1.12 Crawl Error Resolution

GSC → Pages → "Why pages aren't indexed" reviewed monthly

Server errors (5xx) fixed within 24 hours

404s from broken internal links fixed by either restoring pages or updating links

Redirect chains > 1 hop eliminated (use direct 301s, not chained redirects)

Section 2: Site Architecture & Internal Linking (10 Checks)

2.1 Three-Click Depth

Every product page is reachable from the homepage within 3 clicks

Audit orphan products (zero internal links) using Screaming Frog

Add products to multiple relevant categories and collections

Cross-link related products, "customers also bought", "complete the look"

2.2 Category Hierarchy

Categories are siloed: no cross-canonical between unrelated sections

Top-level categories have descriptive, keyword-rich titles (not just "Products")

Subcategories are 1-2 levels deep; avoid 4+ level deep nesting

Empty categories are either populated or 404'd — never left as empty indexable pages

2.3 Breadcrumb Navigation

Every product page renders BreadcrumbList structured data

Breadcrumbs visually match the URL hierarchy exactly

BreadcrumbList schema includes the full path with absolute URLs

The breadcrumb is also visible in the page UI (not just JSON-LD)

2.4 Faceted Navigation Discipline

Either `noindex, follow` on filter combinations (preferred for SEO), or

Use `rel=canonical` to point filters to the unfiltered parent, or

Use `Disallow` in robots.txt for filter-only URLs

Never allow unlimited facet combinations to be indexed (creates bloat)

Prioritize which facets get indexed based on search volume (size, brand, color for apparel)

2.5 Internal Link Anchor Diversity

Use keyword-rich anchors in body content ("blue cotton t-shirt" not "click here")

Avoid over-optimized exact-match anchors (looks manipulative)

Footer and header links use brand-neutral anchors

Product recommendation modules use descriptive anchors

2.6 HTML Sitemap

For sites with 1,000+ SKUs, a paginated HTML sitemap helps users and crawlers

Categories are organized by topic, not alphabetical (better for thematic authority)

HTML sitemap is linked from the footer

HTML sitemap mirrors the XML structure

2.7 404 and 410 Pages

Custom 404 page with search box, popular categories, and featured products

410 (Gone) used for permanently removed products

Soft 404 detection automated via Screaming Frog custom extraction

404 broken-link report exported monthly from ahrefs/Semrush

2.8 Redirect Map

All 301 redirects documented in a redirect map spreadsheet

No redirect chains > 1 hop

No redirect loops

Product retirement uses 301 to closest category page, not just to homepage

2.9 Out-of-Stock Strategy

Out-of-stock products remain live for 30-90 days with "Notify me" CTA

After 90 days, either 301 to a related product or 410 the URL

Never blanket noindex out-of-stock products — you lose accumulated authority

Keep the URL, title, description, images, and reviews — only swap the buy button

2.10 Search Results and Tag Pages

Internal site search results: `noindex, follow`

Tag pages: `noindex, follow` unless they have unique content

Filter combinations: see 2.4

Sort options: never indexable, block in robots.txt

Section 3: Core Web Vitals (8 Checks)

3.1 Largest Contentful Paint (LCP) - Target < 2.5s

LCP element (usually hero image or H1) loads within 2.5s on 75th percentile mobile

Hero images use `<Image priority>` (Next.js) or `fetchpriority="high"`

Images are WebP or AVIF, sized to viewport, with width/height attributes

Critical CSS is inlined, non-critical CSS deferred

CDN serves static assets (Cloudflare, Fastly, Bunny CDN)

3.2 Interaction to Next Paint (INP) - Target < 200ms

INP replaced FID in March 2024 — measure INP, not FID

Main thread is not blocked by long tasks (> 50ms)

Third-party scripts (analytics, chat widgets, ad tags) are deferred or lazy-loaded

Event handlers are debounced/throttled

Use Chrome DevTools → Performance → Web Vitals to measure real user flows

3.3 Cumulative Layout Shift (CLS) - Target < 0.1

Every image and video has explicit `width` and `height` attributes

Web fonts use `font-display: swap` and preloaded

Dynamic content (banners, cookie notices) reserves space before load

Ads and embeds have placeholder dimensions

3.4 Mobile Performance

Mobile LCP and INP tracked separately (mobile is 100% of Google's crawl as of 2024)

Tap targets are at least 48x48px with 8px spacing

Font size minimum 16px base

No horizontal scroll on any viewport 320px+

Lighthouse mobile score > 90 for the homepage and category pages

3.5 Image Optimization

All product images use WebP or AVIF (with JPEG fallback)

Lazy-load below-the-fold images (`loading="lazy"`)

Responsive images use `srcset` with appropriate breakpoints

Image CDN handles on-the-fly resizing (Cloudflare Images, imgix, Cloudinary)

Decoded image dimensions match container size (no oversized images)

3.6 JavaScript Performance

Critical JS inlined or loaded with `defer`

Non-critical JS loaded with `async` or after user interaction

Code-split per route (Next.js does this automatically)

Third-party tags loaded via Tag Manager with appropriate triggers

Bundle size < 200KB initial, < 1MB total

3.7 CSS Performance

Critical CSS inlined in `<head>`

Unused CSS purged (PurgeCSS, Tailwind's content config)

CSS file size < 50KB gzipped

No render-blocking external CSS

3.8 Third-Party Script Audit

Every third-party script (GA4, GTM, Meta Pixel, chat widget, review platform) is justified

Scripts load with appropriate defer/async strategies

Marketing tags fire on conversion events, not page load

Heavy widgets (chat, exit intent) load after 5s or on scroll

Section 4: Structured Data for E-commerce (10 Checks)

4.1 Product Schema

Every product page has Product JSON-LD with: name, image, description, sku, gtin/mpn, brand, offers

Offer block includes: price, priceCurrency, availability, url, priceValidUntil

AggregateRating present if reviews are displayed on the page

Multiple offers (variants) properly nested

4.2 Review Schema

Review schema matches visible on-page reviews

AggregateRating averages match the displayed star count

Reviewer names match the actual customer (no fake reviewer placeholders)

reviewCount matches visible review count

4.3 BreadcrumbList Schema

Every product page has BreadcrumbList matching visible UI breadcrumb

Each ListItem has name, item (absolute URL), and position

Item URLs are canonical and 200-status

4.4 Organization Schema

Organization schema on the homepage with logo, contactPoint, sameAs

Knowledge Graph signals align (Wikipedia, Wikidata, social profiles)

Logo follows Google's logo guidelines (square or wide, no transparency issues)

4.5 WebSite Schema with SearchAction

WebSite schema includes a SearchAction potentialAction

`target.urlTemplate` points to a working internal search URL

`query-input` is properly formatted

4.6 FAQPage Schema

Visible FAQ sections on category or product pages have matching FAQPage schema

Each question in `mainEntity` matches a visible question

Accepted answers contain the full answer text, not truncated versions

4.7 ItemList Schema

Category pages with product lists include ItemList schema

`itemListElement` references product URLs (not category URLs)

Position reflects actual ordering on the page

4.8 LocalBusiness Schema (if applicable)

Physical retail stores add LocalBusiness schema to their contact/location page

Includes address, geo coordinates, openingHoursSpecification, telephone

AggregateRating included if reviews are collected

4.9 VideoObject Schema

Product videos include VideoObject schema with name, description, thumbnailUrl, uploadDate

Content URL is a direct video file, not a YouTube/Vimeo embed

For embedded YouTube videos, use `contentUrl` pointing to the watch URL

4.10 Schema Validation

All structured data passes Google's Rich Results Test

Schema.org validator confirms valid JSON-LD syntax

No deprecated types (HowTo, SpecialAnnouncement, CourseInfo, ClaimReview)

Section 5: On-Page SEO (10 Checks)

5.1 Title Tags

Every product page has a unique title: `[Product Name] - [Spec/Benefit] | [Brand]`

Length 50-60 characters

Primary keyword front-loaded

No keyword stuffing

No duplicate titles across products

5.2 Meta Descriptions

Unique per page, 140-160 characters

Includes primary keyword, a value prop, and a CTA

Matches page content (no clickbait)

Not the same across product variations

5.3 H1 Tags

Exactly one H1 per page

H1 matches or closely aligns with the product name

H1 includes primary keyword naturally

H1 visually styled (no hidden H1s)

5.4 Heading Hierarchy

Logical H1 → H2 → H3 nesting

No skipped levels (H1 → H3 without H2)

Subheadings are descriptive, not generic

5.5 Product Descriptions

300+ words unique per product (no manufacturer copy-paste)

Includes use cases, specifications, sizing info, and FAQs

Naturally incorporates semantic keyword variations

Differentiates from competitor descriptions

5.6 Image Alt Text

Every product image has descriptive alt text

Alt text includes the product name naturally

No keyword stuffing in alt text

Decorative images use `alt=""`

5.7 Internal Linking

3-5 internal links per product page to related products, categories, and content

Anchor text is descriptive, not generic

No orphan product pages

Footer and header link to primary category hubs

5.8 URL Canonicalization

Each product has exactly one canonical URL

Variant pages (color, size) canonicalize to the parent

Trailing slashes consistent (or all redirect to the preferred version)

5.9 Pagination Tags

Either `view-all` canonical or `rel=next/prev`

No-index on deep pagination (page 50+) when no longer useful

5.10 Thin Content Detection

Identify products with < 100 words of unique content

Either expand or noindex

Avoid product pages with only image and price (no description)

Section 6: Duplicate Content (5 Checks)

6.1 Product Variation URLs

Each product has one canonical URL; variations either canonicalize or are noindexed

SKU-based variants use the same canonical as the parent

Size/color swatches are not separate indexable URLs

6.2 Parameter-Based URLs

Sorting parameters (?sort=price_asc) are not indexable

Tracking parameters (?utm_source=) are stripped via canonical

Filter combinations are noindex or canonicalized

6.3 Print-Friendly and AMP Versions

AMP pages (if used) have canonical pointing to the non-AMP version

Print pages noindex

6.4 Staging and Preview URLs

Staging environments blocked via robots.txt and password protected

Preview URLs not linked from anywhere

6.5 Cross-Domain Content

Marketplace listings (Daraz, Amazon) with duplicate descriptions are protected via canonical pointing to your own site

Affiliate URLs use nofollow and `rel="sponsored"`

Section 7: International & Multi-Region (5 Checks)

7.1 Hreflang Implementation

All language/region variants have reciprocal hreflang

`x-default` set to the global or auto-detection page

Self-referencing hreflang on every page

Hreflang in HTML head, not just sitemap

7.2 Currency and Pricing

Schema Offer.priceCurrency matches the target region

Local currency displayed in HTML

Exchange rates not hard-coded (use a dynamic rate)

7.3 ccTLDs vs Subdirectories

Either ccTLDs (`store.bd`, `store.in`) or subdirectories (`/bd/`, `/in/`) — don't mix

Search Console properties configured for each version

7.4 Content Localization

Translations are human-quality, not machine-translated

Local cultural references and examples included

Local units of measurement (cm, kg, BDT)

7.5 Shipping and Tax Information

Local shipping options displayed

Tax/duty information accurate per region

Returns policy localized

Recommended Tools for the Audit

Screaming Frog SEO Spider — Full site crawl, schema extraction, custom extraction

Google Search Console — Index coverage, crawl stats, Core Web Vitals

Google PageSpeed Insights + CrUX — Real user CWV data

Lighthouse (in Chrome DevTools) — Lab performance and accessibility

ahrefs or Semrush — Backlink profile, content gaps, technical health score

Rich Results Test — Schema validation

Schema.org Validator — JSON-LD syntax check

Chrome DevTools → Performance — Real user INP measurement

Server log file analyzer (Screaming Frog Log Analyzer) — Crawl budget analysis

Audit Cadence and Prioritization

Priority
P0 (Critical)
P1 (High)
P2 (Medium)
P3 (Low)

Run a full audit quarterly. Run the P0 and P1 checks monthly. Watch GSC's Page Indexing report weekly for new issues.

From Audit to Action

A technical audit without a follow-through plan is just a report. The next step is prioritization: pick the 3-5 highest-impact fixes and ship them in the first sprint. Then iterate. A technical SEO expert in Bangladesh can run this audit in 2-3 days and ship the first round of fixes within 1-2 weeks.

See how a comprehensive SEO strategy helped Walton Group achieve 53,600+ keyword rankings through technical optimization and semantic content. For a comprehensive approach to search visibility, combine your technical SEO foundation with semantic SEO strategies and learn about how AI is transforming modern SEO. Need hands-on help? Hire an expert to manage your technical SEO. Explore my technical SEO services. For the mobile-specific CWV, mobile-first indexing parity, and 3G/4G-tuned rendering that ship the bottom of your conversion funnel, see my mobile SEO for Bangladesh guide.