Technical SEO Case Study in Pakistan

The Client

A Karachi-based retailer selling consumer electronics and home appliances — smartphones, laptops, televisions, kitchen appliances, and small electronics — through a custom-built ecommerce store. The catalog sat at roughly 3,400 SKUs spread across 60 categories and a few hundred brand and filter combinations. They ran two physical showrooms in Karachi and had built the online store three years earlier as the primary growth channel.

The business worked, but it was increasingly paid-dependent. Google Shopping and Meta ads drove the majority of online revenue, and the team had built a competent paid operation around them. Organic, however, was quietly deteriorating. Over the previous twelve months, category pages — the pages that should rank for high-volume commercial queries like “LED TV prices in Karachi” or “best air fryer Pakistan” — had been steadily dropping out of Google’s index. Where the store had once ranked for dozens of category terms, it now appeared for a shrinking handful, almost all branded.

The leadership team’s concern was straightforward: paid acquisition costs were rising, and they were losing the organic visibility that should have provided a margin cushion. They engaged WeProms Digital for a single-quarter, technical-SEO-first implementation: restore crawl health, get category pages back into the index, and fix the mobile experience that was quietly dragging everything down. This case study is an illustrative composite built from the patterns we see across Pakistani ecommerce catalogs; the figures are realistic outcome ranges for sanity-checking fit, not audited results.

To frame the scale: the store was turning over roughly PKR 90 to 110M a month online, of which paid channels carried an estimated 70%+, organic around 12%, and the remainder from direct, brand, and showroom-referred traffic. A healthy electronics catalog of this size in Pakistan should see organic sit closer to 25 to 35% of sessions given the search demand for category and brand terms, so the gap was not a market problem — it was a discoverability problem. The dev team was small (two in-house engineers plus a freelance frontend contractor), which shaped how we sequenced the work: fixes had to be high-leverage, clearly specified, and prioritised so a thin engineering team could ship them inside a single quarter without stalling other roadmap work. Everything in the plan below was scoped against that constraint.

The Problem

The diagnosis surfaced four technical problems, each compounding the others:

1. Crawl errors and soft-404s. Search Console reported over 1,900 errors — a mix of genuine 404s, server timeouts on deep paginated paths, and a large volume of soft-404s where faceted pages returned empty results with a 200 status. Google was wasting crawl budget on broken and empty URLs.

2. Collapsing category indexation. Only 41% of category pages were indexed, down from 78% a year earlier. The cause was not a single penalty — it was crawl-budget exhaustion. With tens of thousands of low-value duplicate URLs competing for attention, Google was deprioritising and eventually dropping legitimate category pages.

3. Faceted navigation generating exponential duplicates. The store’s filter system (brand, price range, colour, in-stock, specification) generated a new indexable URL for every combination. A crawl of the site surfaced over 50,000 duplicate or near-duplicate URLs built from facet permutations — most of them thin, empty, or identical in content to the parent category.

4. Poor mobile Core Web Vitals. Category pages loaded slowly on mobile, with LCP averaging 4.6 seconds, driven by unoptimised product images, render-blocking scripts from a chat widget and a recommendation engine, and layout shift from dynamically injected elements. Google’s mobile-first indexing meant this was not just a UX issue — it was a ranking issue.

5. Sitemap and crawl-priority neglect. The XML sitemap had not been maintained in over a year and still referenced hundreds of discontinued products and parameter URLs, while the most important category pages were absent from it entirely. Combined with no clear internal-link hierarchy, Google had no reliable signal about which pages the business actually considered important — so it defaulted to its own heuristics, and those heuristics were being fooled by the duplicate-URL flood.

The net effect: Google could not efficiently crawl the site, was dropping the pages that mattered most, and was scoring the mobile experience poorly. Organic was decaying while paid costs rose.

Phase 1 — Technical Crawl Audit and Error Inventory (Weeks 1-2)

We opened with a full technical SEO audit to produce a precise, prioritised error inventory before changing anything. Guessing at technical fixes is how teams spend a quarter without moving the numbers.

Crawl and log analysis. We ran a Screaming Frog crawl, combined it with Google Search Console coverage reports, and — critically — reviewed server log data to see what Googlebot actually requested versus what the site offered. The logs told the real story: Googlebot was spending a large share of its crawl budget on faceted and parameter URLs, many of which returned 200 with empty or duplicate content, and was running out of budget before reaching deeper category and product pages.

Quantifying the log data made the diagnosis undeniable. Of the crawl requests Googlebot made in the sampled window, a disproportionate share hit parameterised and faceted URLs, while the core category and brand pages were crawled infrequently — many not within the last several weeks. Crawl budget is finite and roughly proportional to a site’s perceived health and authority; when the majority of it is consumed by duplicate and empty URLs, the pages that drive revenue simply stop being refreshed. This is why category pages had been silently dropping: they were not being penalised, they were being deprioritised through neglect. Reframing the problem as crawl-budget allocation rather than “ranking” reframed the entire fix.

Error inventory. We categorised the 1,900+ reported issues into actionable buckets:

Error type	Count	Resolution path
Genuine 404 / broken links	~210	301 redirect or proper 404
Soft-404 (empty faceted pages, 200 status)	~740	noindex + parameter handling
Server timeouts on deep pagination	~120	Crawl-budget and pagination fix
Important URLs blocked by robots	~60	robots.txt correction
Duplicate content / canonical conflicts	~770	Canonicalisation and consolidation

Quick wins in week one. Before the larger structural fixes, we deployed the low-risk, high-speed corrections: fixed the 60 important URLs accidentally blocked by robots.txt, implemented proper 301s for the 210 broken links, and corrected a misconfigured canonical that was pointing category pages to the homepage. These alone recovered a handful of category pages into the index within ten days.

Outcome of Phase 1: a documented, prioritised fix list mapped to revenue-impact priority (category and brand pages first), and an early indexation lift from the quick wins.

Phase 2 — Faceted-Nav Control and URL Consolidation (Weeks 2-5)

This was the single highest-leverage phase. The 50,000+ duplicate URLs were not a content problem — they were a crawl-budget problem, and the fix was structural.

Faceted-navigation governance. We implemented a clear rule set for which facets should be crawled and indexed:

• Indexable facets — only the primary, value-creating facets (for example, a category filtered by a major brand) — kept crawlable with self-referencing canonicals and unique metadata.
• noindex, follow applied to secondary facets (colour, price-band, in-stock) that aid user experience but should not compete in the index.
• robots.txt disallow plus URL parameter handling in Search Console for parameter combinations that produced pure duplicates or empty results.

The result: the indexable URL surface contracted sharply and intentionally. After deployment, the duplicate URL count dropped from over 50,000 to under 4,000 — each remaining URL now earned its place in the crawl queue.

Two implementation details determined whether this actually held. First, we made sure that noindexing a facet combination did not orphan the products it filtered — those products remained reachable via the canonical category path and primary facets, so no SKU lost its crawl path. Second, we added self-referencing canonicals to every surviving indexable facet URL and gave each one unique, descriptive metadata, so a category filtered by a major brand became a genuinely useful, distinct page rather than a near-duplicate. The distinction matters: the goal was not the smallest possible index, it was an index where every URL earned its place and contributed something a buyer would actually search for.

Pagination and crawl depth. Deep category pagination had been timing out and wasting budget. We flattened crawl depth so that every category page was reachable within three clicks from the homepage, implemented proper rel next/prev handling, and added XML sitemap entries for the canonical category and brand pages to accelerate discovery.

Canonicalisation and URL consolidation. The 770 duplicate-content conflicts were resolved with a consistent canonical strategy — every product and category accessible via multiple paths pointed to a single canonical URL. We removed internal links pointing to parameterised variants and standardised the internal-link graph toward canonical URLs.

Outcome of Phase 2 (by week 5): crawl errors down roughly 55% from baseline, the duplicate-URL problem largely solved, and Googlebot’s crawl distribution visibly shifting toward category and brand pages in the logs.

Phase 3 — Indexation, Schema, and Core Web Vitals (Weeks 4-9)

With crawl health restored, we turned to the two factors that determine whether category pages rank once they are indexed: structured data and mobile page experience.

Structured-data implementation. We deployed a clean schema markup layer across the catalog:

• Product schema on every product page (price, availability, rating)
• ItemList and CollectionPage schema on category pages, with breadcrumb navigation
• Organization and WebSite schema sitewide, plus a sitelinks searchbox
• FAQ schema on category buying-guide sections

This was not decorative — structured data gave Google clearer entity signals for each category and product, supported rich-result eligibility, and reinforced the canonical page as the definitive version of each topic.

We validated the markup against Google’s Rich Results test as each template shipped, and corrected a set of legacy errors — products marked up with prices in the wrong currency field, availability values that did not match actual stock, and breadcrumb schema that skipped the category level. Clean structured data does not guarantee rankings, but malformed structured data is a reliable way to lose rich-result eligibility and to confuse Google about which page owns a given product entity. On a 3,400-SKU catalog, getting the product and breadcrumb templates right once paid off across every page that used them.

Core Web Vitals — mobile-first. Because the store was mobile-first indexed, the 4.6s LCP was actively suppressing rankings. We worked through the category-page performance budget with the client’s dev team:

• Converted product images to WebP with responsive srcset, and lazy-loaded below-the-fold images — the single biggest LCP win.
• Deferred the chat widget and recommendation-engine scripts that were render-blocking, loading them after first interaction.
• Preloaded the LCP hero image and the primary font.
• Removed a layout-shifting promotional banner that injected after load, stabilising CLS.

Vital	Before	After
LCP (mobile, category pages)	4.6s	1.8s
CLS	0.31	0.06
INP	290ms	150ms

Indexation requests. For high-priority categories still not indexed, we used Search Console’s URL inspection and request-indexing workflow selectively, and kept the XML sitemap tight and accurate so recrawling was efficient rather than noisy.

Outcome of Phase 3 (by week 9): category-page indexation climbed past 60%, Core Web Vitals passing on mobile across category pages, and product rich results beginning to appear in the SERP for high-volume items.

Phase 4 — Monitoring, Recrawl Acceleration, and Validation (Weeks 8-12)

The final phase locked the gains in and built the monitoring to catch regressions — technical SEO wins are easily lost if no one watches them.

Recrawl acceleration. With the sitemap cleaned and canonicals consistent, we monitored log files to confirm Googlebot was now reaching category pages efficiently and frequently. Pages that had been stale for months were re-crawled and re-cached within the quarter.

Regression monitoring. We stood up lightweight monitoring:

• A weekly crawl snapshot flagging any new error spike, broken link, or canonical regression.
• Search Console alerts for coverage drops.
• Core Web Vitals tracking so that any new script or image regression was caught before it affected rankings.

Validation against revenue-relevant pages. Throughout, we prioritised by commercial value — the categories and brands with real search demand got attention first. This is the ecommerce marketing discipline that turns a technical exercise into a revenue one: every fix was ranked by the commercial weight of the page it recovered.

Outcome of Phase 4 (by day 90): crawl errors down 73% from baseline, category indexation at 67% (up from 41%), and non-branded organic clicks up 37% as recovered categories began ranking for commercial queries again.

Final Results at One Quarter (90 Days)

Metric	Before	After	Change
Crawl errors	1,900+	~510	-73%
Category-page indexation	41%	67%	+64%
Duplicate facet URLs	50,000+	<4,000	-92%
Mobile LCP (category pages)	4.6s	1.8s	-61%
Non-branded organic clicks	Baseline	+37%	Day 90
Indexed category pages	25	40	+60%

What Made This Work

1. Log files told the truth Search Console did not. The reported error count understated the real problem: where Googlebot was actually spending its budget. Reviewing logs revealed that faceted URLs were starving category pages of crawl attention — the root cause of the indexation collapse.

2. Structural faceted-nav control, not content production. The fix was not to write more content. It was to contract the indexable URL surface so each remaining page earned crawl budget. Once category pages were reliably crawled and indexed, they ranked on their own merits.

3. Mobile Core Web Vitals were a ranking lever, not just UX. On a mobile-first-indexed store, a 4.6s LCP was not only losing customers — it was suppressing rankings. Getting it to 1.8s compounded with the indexation recovery.

4. Commercial prioritisation throughout. Not all 60 categories mattered equally. Ranking fixes by commercial search demand meant the quarter produced revenue-relevant indexation gains, not vanity crawl metrics.

5. Monitoring turned a one-off fix into a durable gain. Technical SEO is not a set-and-forget channel. New facets get added, scripts get appended by marketing, product images get uploaded unoptimised. The weekly crawl snapshot and Core Web Vitals tracking meant regressions were caught in days, not quarters — which is the difference between a store that holds its indexation and one that slowly decays again.

What Teams Can Apply

For Pakistani ecommerce stores with large catalogs and declining organic visibility:

1. Check what Googlebot actually crawls, not just what errors are reported. Server log review often reveals a crawl-budget problem that Search Console hides.
2. Govern faceted navigation aggressively. If every filter combination creates an indexable URL, your catalog is fighting itself. Decide which facets deserve indexation and noindex the rest.
3. Tighten crawl depth and sitemaps. Every important category and brand page should be reachable in three clicks and present in a clean XML sitemap.
4. Treat Core Web Vitals as a ranking input on mobile. Image format, render-blocking scripts, and layout shift are the usual suspects — fix them in that order.
5. Prioritise by commercial demand. Fix the categories with real search volume first; that is where recovered indexation turns into revenue.

WeProms Digital has applied this technical SEO framework across Pakistani ecommerce stores in electronics, fashion, grocery, furniture, and home appliances, on platforms from custom builds to WooCommerce and Magento. The specific stack and catalog shape change; the crawl-first, indexation-focused, mobile-vitals-driven approach stays consistent.