Client results

Case Studies: traffic, leads, sales, Maps and AI Overviews

Real SEO projects, what was wrong, what I changed and what it delivered. From 10x organic traffic to tripled sales conversions and a place in the Google Local 3-Pack.

10x organic traffic in one month
+180% organic visibility within weeks
3x sales conversions
5x local organic leads

Prefer Saad Raza SEO on Google Add this site as a preferred source and Google will show more of my SEO and AI search guides when you search.

10x organic traffic in one month Traffic growth · Lead generation · Established business website 10x organic traffic in one month with a holistic SEO overhaul Read the case study
+180% organic visibility within weeks AI Overviews · Traffic growth · Brand website Into Google's AI-generated results with schema and semantic content Read the case study
3x sales conversions Sales growth · Traffic growth · E-commerce store Rich results that tripled sales conversions for an online store Read the case study
5x local organic leads Google Maps · Lead generation · Regional service business From invisible to the Google Local 3-Pack: 5x local leads Read the case study
+35% click-through rate Traffic growth · Content website Winning Google's carousel for competitive top 10 searches Read the case study

Traffic growthLead generation Established business website

10x organic traffic in one month with a holistic SEO overhaul

The challenge

The client had been stuck at the same traffic level for months, even though new content went live regularly. Quantity was never the problem: the site had technical errors, weak authority and keyword targeting scattered across overlapping pages.

What I did

Ran a technical audit and fixed Core Web Vitals, broken links, crawl errors and indexing issues.
Ran a content gap analysis and reorganised the site into topic clusters, each with a pillar page and supporting articles.
Rewrote title tags, meta descriptions and heading structures around clear search intent.
Built an internal linking plan so authority flowed to the pages that matter most.
Earned high-quality backlinks from relevant industry sites and pushed new content through the brand's distribution channels.

AI OverviewsTraffic growth Brand website

Into Google's AI-generated results with schema and semantic content

The challenge

The client wanted the brand to appear in Google's AI-generated search results, not just the classic rankings. The content was good, but the site lacked structure, semantic depth and proper schema, so search engines were not interpreting it correctly.

What I did

Implemented structured data across the site, including Article, Organization, FAQ and HowTo where they applied, so Google could identify entities and how they relate.
Refined the content with NLP principles: clearer entity usage, natural keyword variations and coverage of the questions behind each search.
Restructured titles, meta descriptions and H1 to H2 hierarchies to match search intent.
Fixed navigation and mobile usability issues that were hurting the experience.

Sales growthTraffic growth E-commerce store

Rich results that tripled sales conversions for an online store

The challenge

The store ranked reasonably well for product keywords but was not getting the clicks. Its listings had no rich snippets, and many product pages suffered from duplicate titles, descriptions and content.

What I did

Fixed duplicate title tags and meta descriptions and cleaned up the URL structure for efficient crawling.
Implemented Product schema with ratings, reviews, price and availability.
Kept product data consistent with Google Merchant Center.
Rewrote product descriptions around buyer-intent keywords.
Strengthened calls to action and added comparison content to help shoppers decide.

Google MapsLead generation Regional service business

From invisible to the Google Local 3-Pack: 5x local leads

The challenge

A regional service business was struggling against larger national brands. It did not appear in Google Maps or the Local 3-Pack, where most of its customers were choosing who to call.

What I did

Optimised the Google Business Profile: services, categories, description and consistent name, address and phone number everywhere.
Created geo-targeted landing pages for the key service areas.
Optimised for near me and location-based searches.
Built a process to request genuine reviews from real clients.
Listed the business on high-authority local directories with consistent details.

Traffic growth Content website

Winning Google's carousel for competitive top 10 searches

The challenge

A content website wanted to rank for highly competitive list queries such as top 10 tools and best platforms. It had the content, but it was not structured or optimised for Google's carousel results.

What I did

Rebuilt the article as a structured, scannable and semantically rich top 10 list.
Added list structured data to signal eligibility for carousel results.
Expanded depth with long-tail variations, FAQs, comparisons and alternatives.
Added visuals, tables and bullet points, and used evergreen angles that suit Google Discover.

How I deliver results like these

Baseline firstSearch Console, analytics, Maps rankings and a fixed AI prompt set are recorded before anything changes.
Fix what holds you backTechnical, content and authority problems are fixed in order of impact, not all at once and not at random.
Report what movedMonthly reports show traffic, leads, sales and visibility against the baseline, with what comes next.

How I work

AI search and AI agents: my playbooks

The methods behind my AI Overviews, LLM and agentic SEO work, step by step.

01 Live project Professional services, Karachi Rebuilding saadrazaseo.com for AI answers and AI agents My own site, rebuilt in the open: a new theme, 124 city and industry pages, 33 cluster articles and one connected schema graph.

The starting point

The site had grown into a mix of page builders, three overlapping SEO and schema plugins, location pages generated from one template, and a homepage form that did not send anywhere. AI systems and agents had no clean, consistent picture of who Saad Raza is, what he offers or where he works.

What I did

Replaced the page-builder header, footer and homepage with a lightweight custom theme, and removed the leftovers that caused layout shift.
Rewrote all 124 city and industry pages so each one answers questions specific to that market, instead of swapping the city name into one template.
Published 33 articles in nine topic clusters, each linked to its service page and to the other articles in its cluster.
Moved all SEO data into one plugin and built a single connected JSON-LD graph: Person, ProfessionalService, Service, BlogPosting, BreadcrumbList and FAQPage.
Self-hosted fonts with metric-matched fallbacks, so text no longer moves when the web font loads.
Fixed the lead form so it submits on the front end, without being blocked by the host security challenge that also stops AI agents.

02 AI search visibility B2B SaaS How I earn citations in AI Overviews and AI Mode for SaaS brands Strong blog traffic, but the brand is missing when buyers ask AI tools to compare products in its category.

The typical problem

A SaaS company ranks for many informational keywords, yet AI Overviews and chat assistants cite review sites and competitors when buyers ask comparison and how-to questions. Its feature pages are thin, pricing sits in images and the comparison content is gated.

How I solve it

Build a prompt set from sales calls and Search Console: the comparison, alternative and how-to questions buyers really ask.
Check each prompt in AI Overviews, AI Mode, ChatGPT, Gemini and Perplexity, and record which sources get cited and why.
Rewrite key pages answer-first: a direct answer in the first paragraph, then specifics, tables and limits.
Move pricing, integrations and specifications out of images and gated PDFs into crawlable HTML.
Publish original material others want to cite, such as benchmarks or anonymised product data, with a named author and method.
Strengthen the entity: consistent Organization and Product markup, author pages and coverage on the third-party sites AI tools already cite.

03 AI search visibility Private healthcare, Karachi How I get local clinics recommended by AI assistants Patients ask ChatGPT and Gemini for the best clinic near them, and the answers name competitors.

The typical problem

A well-reviewed private clinic appears in the local pack but rarely in AI assistant answers. Its name, address and phone number differ across directories, doctor profiles have no credentials, and treatment pages are generic.

How I solve it

Audit how the clinic is described across Google Business Profile, directories, social profiles and health portals, then fix every mismatch.
Create doctor profile pages with qualifications, registration details and specialties, marked up as Physician and linked to the clinic.
Rewrite treatment pages around the questions patients ask: cost factors, preparation, recovery and when to see a doctor, reviewed by the clinic doctors.
Build a review process that asks patients to mention the treatment they had, without incentives, staying within Google review policies.
Track a fixed set of local prompts monthly across the main assistants.

04 Agent-ready sites Home services, multi-city How I make booking flows that AI agents can complete Customers ask browser agents to book a service, and the agent gives up halfway through the form.

The typical problem

A home services company gets traffic from AI browsing agents, such as ChatGPT agent mode and agent features in Chrome and other browsers, but bookings fail. The form uses icon-only buttons and placeholder text instead of labels. Prices load after a delay that shifts the layout, and a bot challenge appears before the first step.

How I solve it

Run the booking task with several browser agents and record exactly where each one fails.
Give every field a visible label and an accessible name, and replace icon-only buttons with clear text such as Book a visit.
Render prices, service areas and time slots in the HTML instead of injecting them late, so the layout does not move.
Move bot protection to the final submission step and use rate limits instead of a challenge on page load.
Write plain-language policies on cancellations, call-out fees and coverage that an agent can read and repeat accurately.

05 Agent-ready sites Fashion e-commerce How I prepare online stores for AI shopping agents AI shopping features recommend products from feeds and structured data, and this store sends incomplete data.

The typical problem

A fashion retailer wants its products to appear when shoppers ask AI assistants for recommendations. Its Merchant Center feed lacks size, colour and material attributes, variants share one URL, stock status lags behind the website and the returns policy is only in a PDF.

How I solve it

Audit the product feed and on-page Product markup against Google merchant listing requirements and the attributes AI shopping features use.
Give each variant its own URL or identifier, with size, colour, material, GTIN where available and accurate stock status.
Publish shipping and returns policies as HTML pages and in structured data.
Sync price and availability between the store, the feed and the markup so agents never see stale data.
Rewrite product descriptions with the facts shoppers and agents compare: fit, fabric, care and sizing advice.

06 AI search visibility Online publisher How I set an AI crawler policy for publishers The publisher blocked every AI bot after a server overload, and its articles disappeared from AI answers.

The typical problem

After a traffic spike, a publisher blocked all AI user agents in robots.txt and its firewall. That also blocked the search and user-triggered fetchers that let AI tools cite and link to its articles, while some training crawlers still got through via unblocked paths.

How I solve it

Analyse server logs to separate training crawlers from search and user-triggered fetchers, such as GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, PerplexityBot and Google-Extended.
Verify bots by reverse DNS or published IP ranges, so fake user agents are not trusted.
Set a policy that allows search and citation fetchers, decides on training crawlers as a business choice and rate-limits rather than blocks where load is the problem.
Remember that Google-Extended controls Gemini model use, not Google Search or AI Overviews, so do not block Googlebot.
Add an llms.txt file as a low-cost experiment, without expecting ranking benefits: it is a proposed convention, not a search engine standard.

07 AI-assisted SEO operations Content site, 2,000+ articles How I use AI agents to find and fix content decay Too many articles to review by hand, so outdated pages quietly lose clicks every month.

The typical problem

A large content site knows many of its articles are out of date but cannot review them all. Past attempts to mass-rewrite pages with AI produced generic text and some pages lost visibility.

How I solve it

Connect Search Console data to an agent workflow that flags pages with falling clicks, ageing facts or overlapping topics.
Have the agent draft a brief for each flagged page: what changed, what is missing and which pages compete with it, not a full rewrite.
Send every brief to a human editor who decides to update, merge, redirect or leave the page, and who writes or approves the final text.
Log every change with a date so results can be compared against pages that were not changed.
Keep the workflow within Google guidance on scaled content: AI helps with triage, people stay responsible for what gets published.

08 AI-assisted SEO operations Affiliate and review site How I clean up mass-produced AI content after a core update Hundreds of AI-written pages went live in a few months, then search visibility fell sharply.

The typical problem

A review site published hundreds of near-identical AI articles. After a core update, visibility dropped across the whole domain, including the original hand-written reviews that used to perform well.

How I solve it

Classify every URL by origin, traffic, links and uniqueness to find the pages that add little value.
Remove or noindex pages with no unique value, merge overlapping ones and redirect where a better page exists.
Rebuild the best pages around first-hand testing: photos, measurements, clear criteria and named reviewers.
Add author pages and an editorial policy that explains how AI is and is not used.
Set a slower publishing process with a human check on every page before it goes live.

09 AI search visibility Manufacturer and exporter, Pakistan to GCC How I get exporters into Arabic AI answers Gulf procurement teams ask AI assistants in Arabic, and the exporter only has English pages.

The typical problem

A Pakistani manufacturer sells to buyers in the UAE and Saudi Arabia. Buyers research suppliers in Arabic and English, but the site has machine-translated Arabic pages with no hreflang, and certifications are hidden in image files.

How I solve it

Research how Gulf buyers phrase supplier and product questions in Arabic and English, instead of translating English keywords.
Have native speakers write the Arabic pages, set up hreflang correctly and give each language its own URLs.
Publish certifications, minimum order quantities, lead times and export markets as crawlable text.
Mark up the manufacturer with Organization and Product data and keep its details consistent on trade directories and platforms.
Test buyer prompts in both languages across the main AI assistants each month.

Questions about these case studies

How long did these results take?

It varies with the starting point and the competition. In these projects the first big movements came within weeks for the AI visibility work, within one month for the holistic traffic project and within three months for the e-commerce store. I give every client a realistic timeline after the audit, never a guarantee.

What is agentic SEO?

Agentic SEO has two sides. The first is making your site easy for AI agents to use: agents that browse, compare and book on a person’s behalf need clear labels, stable layouts, readable policies and forms they can complete. The second is using AI agents inside SEO work, for example to flag decaying pages or draft briefs, with a person reviewing every change before it goes live.

Can you guarantee my business will appear in AI Overviews or ChatGPT answers?

No. Nobody can guarantee placement in AI answers, just as nobody can guarantee rankings. What I can do is remove the reasons you are being left out: thin or gated content, inconsistent business details, missing structured data and pages that crawlers or agents cannot read. Then I measure what changes against a fixed set of prompts.

How do you measure visibility in AI assistants?

I build a fixed set of prompts your buyers actually use and check them on a schedule across AI Overviews, AI Mode, ChatGPT, Gemini and Perplexity, recording mentions, citations and which pages are cited. That sits alongside Search Console, analytics referrals from AI tools and the enquiries you receive.

Want to know where your site stands with AI search and AI agents? Request a free SEO audit and I will send clear, prioritised next steps.