0 to 30,000+ combined monthly organic sessions, built on ten clinic websites over 18 months of one engagement.
Source: that 18 month healthcare network engagement, led by Muhammad Waqas.

A business in Pakistan hires me to win in two places at once: the Google result it already competes for, and the answer an assistant writes when a buyer asks about its category.
I set the strategy and own the account, the specialist team I lead ships the work, and no figure in a report of mine comes from a dashboard you cannot open yourself.
Below is the whole engagement: four pillars of delivery, 57 named items of coverage, the call window mapped to the hour in Pakistan Standard Time, how the work reaches the markets you sell into, three results, each with its period and its source, and my own measurement of the 28 pages that rank in this market.
Last updated November 2026
0 to 30,000+ combined monthly organic sessions, built on ten clinic websites over 18 months of one engagement.
Source: that 18 month healthcare network engagement, led by Muhammad Waqas.
Every figure on this page can be traced. An outside number carries its publisher, its month and the limit the publisher states; a client number carries the engagement it came from and how long that engagement ran.
Free 30 minute strategy call on Google Meet, Monday to Friday, US Eastern business hours. In Pakistan Standard Time that is an evening slot, mapped to the hour further down this page.
Three measured changes are moving this decision in Pakistan right now: an AI answer now sits above the result you used to rank for, the AI positioning reached this market before the search demand did, and most of what ranks for these phrases is a ranked list of other people.
Each one is dated and sourced below.
1. An AI answer now sits above the result you used to rank for.
In the study that measured what people do next, a traditional result was clicked on 8% of visits where a summary appeared, against 15% where none did, and 18% of the searches studied produced a summary at all (Pew Research Center, 22 July 2025, browsing data from 900 US adults and 68,879 searches). Read that as an association on Google and as a US sample, because that is the population Pew measured. The same surface now sits over Pakistani results, which is why this work covers the ranking and the answer together.
2. The positioning arrived in this market before the search box did.
Every AI qualified Pakistan phrase I pulled returns no measurable volume at all, best ai seo consultant in pakistan, ai seo consultant pakistan and ai seo expert pakistan among them, while the consultant and expert phrasing carries 250 searches a month across nine phrases at difficulty 14 to 31 (DataForSEO Labs, read 11 November 2026, country level estimates for the phrase rather than a panel of searchers). Buyers here still type expert. They judge you on the AI work after they land.
3. Most of what ranks for those phrases is a list of other people.
Across four Google reads on 11 November 2026 for these phrases, 12 of the 35 distinct results are ranked lists, counted by their own titles.
By the time somebody opens this page they have usually seen a name on one of those lists and come to check it, which is why the first screen carries the record rather than a pitch.
Five situations bring a business here, and the first move differs in each:
The marketing lead who has been asked for an AI visibility number by a board.
First move: agree the prompt set, then read it, so the first slide upstairs is a measurement.
A group whose branch or location pages split their own demand.
First move: the coverage matrix and an internal link map that gives each page its own queries.
A site that migrated recently, or one whose clicks fell after an AI answer appeared over its result.
First move: the crawl and access read, because reachability explains it more often than content does.
A business whose last supplier sent rankings and no inquiries.
First move: rebuild the reporting on properties you own and on conversions.
A business whose buyers sit in markets other than the one it sells from.
First move: read the entity and the hreflang layer across every market you sell in.
I deliver four pillars to a business in Pakistan: AI visibility and citation, technical SEO and crawl health, semantic content and entity coverage, and measurement you own.
Each pillar below says what gets done and what you hold at the end of it.
This is the work on my AI SEO services page delivered to one market as a senior engagement rather than a seat on a roster. I am Muhammad Waqas, AI SEO consultant, and yours takes its shape on the free call once I have read your site.
What those four pillars are measured on:
Presence in the answers your fixed prompt set returns, prompt by prompt, month by month.
Impressions and clicks on the queries you sell on, read inside the Search Console property you own.
Conversions in your own GA4 property, traced back to the pages this work touched.
The share of your pages an engine is allowed to read, and the share that answer in their first sentence.
Crawl budget spent on the pages you want quoted rather than on parameter and version URLs.
For the wider program this engagement sits inside, see the full SEO service list.
Fifty seven named items sit inside this engagement, grouped in the four pillars the work runs in, and the four types of SEO a buyer asks about all live here: technical, on page, content and off page, with the AI answer layer running across all four.
Every line below is work carried out on your own site, not a category standing in for work. The same fifty seven items read at service level on my AI SEO services page.
Fifteen items: Google AI Overviews, Google AI Mode, ChatGPT, Google Gemini, Perplexity, Claude, Microsoft Copilot, generative engine optimization, answer engine optimization, LLM SEO, a fixed prompt set, a baseline citation read, a monthly re read of the same prompts, citation presence scored prompt by prompt, and the brand mentions and digital PR aimed at citations.
write the prompt set before any work starts and keep it fixed. A prompt set that changes every month measures the prompts instead of the progress.
Seventeen items: crawl budget, indexation, index bloat, parameter and version URLs, a complete redirect map, canonical tags, Core Web Vitals, log file comparison before and after, site migration, hijacked page recovery, internal redirect chains, robots rules read agent by agent, GPTBot access, PerplexityBot access, ClaudeBot access, Google-Extended access, and a curated llms.txt.
That file states what your business is, which pages answer which question, and nothing else, written for your site rather than generated from your sitemap.
The record behind this group is the career rather than one example, and the results below carry it with the period and the source on each one.
open your own robots.txt and search it for GPTBot. Whatever it says, you will know in ten seconds whether that question has ever been answered on your site.
Fourteen items: one H1 per page, question led H2s, answer first sections, answer capsules of 40 to 60 words, a table wherever a set of facts can be one, an FAQ with FAQPage schema, service pages, product and category pages, a topical map, entity clarity in body copy, internal links that follow that map, Service, BreadcrumbList and Organization markup shipped, your business named identically everywhere it appears, and the build shipped on WordPress or on whatever platform your site already uses.
give one page one question. Two pages chasing the same phrase split their own clicks, and on a site with plenty of content that is the most common thing I find.
Eleven items: your own Google Search Console property, your own GA4 property, Search Console query reports, Search Console page reports, GA4 conversions, named prompts you can run yourself in ChatGPT, Perplexity and Google, month over month comparison, a monthly report, a monthly call with the consultant who did the work, DataForSEO estimates labeled as estimates, and the stated standard of what is never counted as a result.
agree in writing, before the first invoice, which numbers count as a result. That conversation is cheap now and expensive in month five.
Getting a Pakistani business named inside an AI answer is three jobs: make the facts quotable, open the access the assistants read through, and weight the plan to the engines this market actually uses.
It runs as a service, and each of the three ends in something you hold.
Generative engine optimization, GEO, is the work of becoming a source a generated answer is built from and names, and the deeper service side of it is my generative engine optimization services page. Answer engine optimization, AEO, is the narrower job of supplying the direct answer to a single question. Both are spelled out here on first use because the short forms are not yet how buyers speak.
The Pakistan field already sells AI search: the named engines and the spelled out vocabulary are on these pages already, as the census further down this page counts. So the vocabulary is not the difference. Published method, named reporting, sourced numbers and the record set out further down this page are.
An engine cannot name a business it cannot resolve, so the entity and the schema come before the writing. I state your brand, your services and your proof as facts rather than claims, in body copy and in markup, then make each section's own first sentence liftable.
Deliverable: the entity and schema baseline as a document, the markup shipped, and an answer capsule under every heading on the pages that matter.
The first thing I check is whether the engines can reach your pages at all. Of the 28 pages ranking in this market on 11 November 2026, 5 name an AI crawler in their robots.txt and 2 disallow the named AI crawlers from the whole site, so on 23 of them the question has never been answered either way. Your own file gets read agent by agent in week one, and you get the result line by line.
I then publish the curated llms.txt named in the coverage list above and confirm Bing indexing, because Microsoft Copilot reads a different index than Google does.
Deliverable: the access read agent by agent, the fixes applied, and the file written for your site.
In Pakistan in October 2026, ChatGPT held 58.92% of AI chatbot use, Google Gemini 32.35%, Claude 4.09%, Perplexity 3.85%, Microsoft Copilot 0.77% and DeepSeek 0.02%, and Google held 94.88% of search. One assistant decides this market, so your plan is weighted to ChatGPT first and then to Google's own AI surfaces, which sit on top of that 94.88%. A plan written for another market gets this order wrong.
Source: StatCounter Global Stats, Pakistan, October 2026, read 11 November 2026. These are shares of measured traffic rather than a survey of people, and StatCounter states no sample and no methodology.
| Engine | Where its answer comes from | What you receive on it |
|---|---|---|
| ChatGPT | Its own retrieval plus the pages it is allowed to crawl | The prompt set read first here, every month, scored prompt by prompt |
| Google AI Overviews and AI Mode | The index that already holds your rankings | Answer capsules on the pages that already rank, so the quote comes from you |
| Google Gemini | Google's index and its own grounding | The same capsule work, checked separately because the answers differ |
| Perplexity | Live retrieval with visible citations | The citation check you can run yourself, with the prompt written out |
| Microsoft Copilot | The Bing index | Bing indexing confirmed, which is a different job from Google indexing |
| Claude | Retrieval plus what the open web says about you | The brand mention work, because this one rewards being described consistently |
The AI engine mix this market's plan is built on is not the American one, and that weighting is the reason this section is specific to your market rather than a template. You get a prompt set written with you for ChatGPT, Perplexity and Google, a first read of what each one says about your category today, and the identical prompts read again at the same point every month. Behind those citations sits an 11 stage AI content pipeline with anti pattern detection and automated quality scoring, held to an 8 out of 10 gate across six metrics on 850+ pages. Source: projects led by Muhammad Waqas.
Deliverable: the weighted plan and the prompt set agreed in writing, with the first baseline read inside month one.
If you sell outside Pakistan, the plan has to reach the markets you sell to, not only the one you sit in.
In this market the selling is already there and the plumbing underneath it is not, which is the gap this section closes and the one measurement on this page that belongs to it alone.
Of the 28 pages ranking in this market on 11 November 2026, 7 sell international SEO by name, 16 use the word global and 12 claim worldwide reach or a count of countries, while 1 carries a single hreflang tag and that page is the only one in the set that is not a Pakistani supplier.
One entity, one description, every market. An engine that meets two different versions of your company resolves you to two weaker things instead of one strong one, so the name, the service facts and the proof are stated identically on the site, in schema and in every third party profile that mentions you. Answer coverage is then written for each market's own questions rather than translated once, because a buyer in another country asks about delivery, terms and support in words your home market never uses.
Deliverable: the entity baseline, the list of profiles that have to agree with your site, and a ranked question set per market.
The layer is three things: hreflang with return tags and x-default, a self referencing canonical per market, and a sitemap that matches what you actually want crawled. That layer is what tells an engine your market versions are one business rather than several or duplicates, and it is also what a careless migration destroys first. The record behind it is migrations of large websites running into thousands of pages, without lost equity. If the United States is one of the markets you sell into, that market has a page of its own: AI SEO consultant in the USA.
Deliverable: the market structure specified, shipped and checked crawler side.
An engagement runs as one senior relationship on your clock: I own the account and the order of work, a specialist on the team I lead ships it, and calls land in a Pakistani evening.
The hours, the ownership and the languages are all below, stated once and concretely.
| Half of the US year | The call window, US Eastern | The same window in Pakistan Standard Time |
|---|---|---|
| While the United States is on standard time | 08:00 to 16:00, Monday to Friday | 18:00 to 02:00 |
| While the United States is on daylight time | 08:00 to 16:00, Monday to Friday | 17:00 to 01:00 |
The Pakistan call window is in that table, and so is the reason there are two rows. Calls run through the Monday to Friday US Eastern business day, which in Pakistan Standard Time is 18:00 to 02:00 while the United States is on standard time and 17:00 to 01:00 while it is on daylight time. The first three hours of that window are an ordinary Pakistani evening, after your working day. Pakistan keeps one clock all year at UTC+5, so the window shifts by an hour twice a year because the American clock moves, never because yours does. The last slots of the week start after midnight, so a Friday call is early Saturday for you.
I own the account, the strategy and the decision about what gets done first, and I am on every call. A specialist on the team I lead executes inside that plan, so the team scale behind this engagement is a fact about capacity. You get one named person who knows your site, the documents from every stage, and the engagement shapes set out on my SEO consulting services page. I speak English (full professional), Urdu (native), Arabic (working knowledge), and the work itself is delivered in English.
The work runs in five steps and every one ends in a document you keep: the free call, the first 30 days, the build, the monthly rhythm, then month 4 onward.
Nothing waits on a signature to become specific, and step 1 produces findings about your own site.
Thirty minutes on Google Meet, inside the evening window above. I open your own robots.txt live and tell you whether the assistants are allowed to read you, which of your pages answer in their first sentence, and the first entity gap I can see.
Deliverable: those three findings in writing, and the scope by the end of the week.
Month one produces the facts that decide the order of everything after it: the entity and schema baseline, the AI crawler access fix list applied, the AI visibility baseline across your fixed prompt set, and the content coverage matrix.
Deliverable: four documents that stay yours, and a publishing order you have signed off.
Answer shaped rewrites on the pages that already rank, technical fixes shipped, the first new pages live, and the market structure put in place where you sell abroad. The order comes out of the coverage matrix rather than a template, and you sign off each item before it ships.
Deliverable: the rewrites and the new pages, each one approved by you.
Every month: the same prompts read again, a report naming what moved and what it was worth, and a call with the consultant who did the work. Every figure comes from your own properties, so you can open any line at source while we are talking.
Deliverable: that month's report, and the call that goes through it with you.
From here the compounding work carries the account: wider coverage, brand mentions and citations, and the crawl or migration work the site needs. Among search specialists, 28% advise a lead time of at least three to four months (Fiverr index via Search Engine Journal, 22 September 2026, a freelancer survey whose total sample and field dates were not disclosed), which is exactly what the four stages above produce.
Deliverable: the commercial queries held month after month, and the revenue the report can trace to them.
Four industries, and the record behind each one is real: healthcare and clinic groups, restaurant and food brands, ecommerce and parking management platforms, and large sites where migration and crawl budget decide the outcome.
I take the work where the proof already exists rather than listing every sector a buyer might sit in.
Clinic groups with several branches are the deepest part of the record: treatment pages written to answer a question rather than describe a service, and branch sites structured so they stop bidding against each other in the same results. No patient facing page carries a health claim I cannot source, and no form on any engagement I run asks for health information.
Food, restaurant and catering websites grown, which here means menu, location and booking pages that answer the question a buyer puts to an assistant, plus the entity work that keeps one brand from reading as several. The same discipline as the clinic record, on a much shorter buying cycle.
Ecommerce and parking management websites grown, where the work is category and facet architecture, index bloat, and product data stated so an engine can quote a specification rather than guess it. On a catalog the entity work decides whether an assistant can name one of your products or only your brand, and the facet and parameter rules decide whether an engine ever reaches the pages that convert at all.
Migrations of large websites running into thousands of pages, and crawl work on sites where an engine was spending its visits on parameter and version URLs. On a site that size, what to write is the second question. The first is which pages an engine reaches at all, which is the point at which a technical SEO consultant earns the engagement.
Three results from engagements I led, every one of them printing its scope, the work, the period and a source line you can check against your own reporting.
That is the standard worth asking any consultant to meet before you weigh a single number.
A result with no scope and no source is a screenshot. Each card below names the engagement the figure came from, how long that engagement ran and who measured it, so you can weigh it instead of taking it.
a single engagement covering all ten websites in the group, over 18 months.
network wide SEO and GEO strategy and execution, treatment page architecture, and an entity and schema layer that lets ten separate sites read as one brand.
over $100M in attributed client revenue across the portfolio, standing on the organic growth in the headline figure.
Source: the 18 month network engagement, led by Muhammad Waqas.
crawl work carried out on more than 20 websites, built up account by account rather than in one project.
parameter and version URLs cleared, indexation repaired and custom schema written. On the clinic site in this example, CSS and JS version parameters were consuming 54% of crawl budget, and all of it was recovered.
engines spending their crawl on the pages the client wanted found, site after site.
Source: those crawl engagements, led by Muhammad Waqas.
migrations of websites whose page count runs into the thousands.
a state level to city level URL directory rebuilt behind a complete redirect map and shipped in a single cutover.
migrated with its equity intact, which is the only outcome worth accepting at that size.
Source: the migration engagements, led by Muhammad Waqas.
The longer write ups behind these three sit in my AI SEO case studies.
Not one figure above is printed without the engagement and the period behind it.
The reason to work with me is four specifics you can check yourself: ten years in search, 80+ specialists led across twelve teams and 20+ concurrent projects, a reporting standard put in writing before the first invoice, and four named certifications.
Not one claim on this page rests on anything but a dated engagement or a named publisher.
The career behind that decade is the whole of it rather than one example: clinic and healthcare groups grown to their maximum organic traffic and conversions; food, restaurant and catering brands grown; parking management platforms and ecommerce stores grown; migrations of sites running into thousands of pages; around 50 hijacked websites recovered.
That team scale is the capacity standing behind one senior engagement: the strategy and the account are mine, and the delivery has a bench behind it. The career behind both facts is set out about Muhammad Waqas.
A reporting standard, put in writing before the first invoice. Every figure comes out of Google Search Console and GA4 properties you own, AI visibility arrives as named prompts you can run yourself, and domain rating, domain authority and brand traffic are kept out of the results entirely. The same standard applies to every service in the full SEO service list, not just to this engagement.
Certifications, exactly four: HubSpot Inbound Marketing, HubSpot Content Marketing, Google Analytics Certified and SEMrush SEO Toolkit.
I measured this field instead of describing it.
On 11 November 2026 I fetched all 28 pages ranking across four Google reads for the consultant and expert phrases in this market and put every one through the same checks. Of those 28 pages:
10 name ChatGPT, 6 name Google AI Overviews and 7 spell out generative engine optimization , so the vocabulary is already here.
1 names the tool a result was measured in beside a result figure , and it does that for one of its three result blocks.
16 serve a usable llms.txt file and 0 mention the file anywhere in their copy. 9 of the 16 name the tool that generated them, the smallest is 237 bytes and lists a single WordPress demo post, the largest is 1.8 MB of links, and 4 carry no description of the business at all.
4 carry any table and 2 carry a booked call link.
That gap is the standard this page is built to. Every result figure on it carries its source in the same card, every outside figure carries its publisher, its month and its caveat, and the llms.txt I write for you says what your business is and which page answers what, rather than listing every URL a plugin can find.
Method: four Google reads on 11 November 2026, all 28 pages fetched on one method the same day, every page saved and every count re runnable. These figures describe 28 pages on one day. They are not a claim about this industry or about any company in it.
A business buying this work has three real options, and they differ on who decides what gets done, who explains the results to you, and what stays with you at the end.
The table compares what you get in each, row by row, as what each shape is good at.
| What you are buying | An agency | A marketplace freelancer | A senior consultant |
|---|---|---|---|
| Who decides what gets done first | A planning team sets the order | You do, task by task | I do, and I am the one who explains why |
| Who explains a month's results to you | The account manager assigned to you | Whoever ran the task | The consultant who decided and did the work |
| The bench behind the work | A roster the supplier staffs | One person | A specialist team I lead and direct |
| AI visibility in scope | An added line on the proposal | A single task | A fixed prompt set, a baseline and a monthly re read |
| Buyers in markets other than yours | Usually a translation request | Usually out of scope | One entity across markets, with the hreflang and canonical layer underneath |
| Where a number in the report comes from | A dashboard the supplier owns | A screenshot | Properties you own and can open while we talk |
| What gets called a result | A visibility or health score | A ranking position | Impressions, clicks, conversions and citations, agreed before the first invoice |
| What stays with you at the end | The reports | The files for that task | Every document, the markup, the redirect map and the prompt set |
That third column is the engagement this page sells, and the shapes it comes in are set out on my SEO consulting services page. What you are choosing there is seniority and ownership, not a headcount.
One named consultant owns the account, and every figure comes out of a property you own.
Ten questions about hiring an AI SEO consultant in Pakistan, every one taken from this market's own searches, from the questions the pages ranking in this field answer, or from what buyers raise on a first call.
Each answer is complete in its first sentence, and each sits in the page markup so an engine can quote it.
Judge it on the record rather than the claim, and mine is on this page in a form you can check: three results each with the scope, the work, the period and a source line, four named certifications, a reporting standard put in writing before the first invoice, and my own measurement of this market published with the method beside it. Ask any candidate for the same four things in the same form, and compare what comes back.
Yes, and it runs as one senior engagement: I set the strategy and the order of work, I am on every call, and a specialist on the team I lead ships the work inside that plan. You hold the documents from every stage, the markup, the redirect map and a report built out of your own properties.
Scope and cost are settled together on the free 30 minute call, after I have read your site, the markets you sell into and the state of your access and entity layer. What moves it is the number of sites, the number of markets, and whether month one is a migration or a build. Engagements are invoiced in US dollars, and you leave that call with the scope in writing.
Yes, and it is consulting rather than a course: your own people do the work on your own site, I set the standards they work to and review everything they ship. It reaches answer shaped page structure, entity and schema practice, crawler access and the monthly prompt read. The capability is still there after the engagement closes, because your own team built it.
Month 1 buys the baseline and the applied fix list, months 2 and 3 bring the first movement and the first AI citations, months 4 to 6 move head terms, and from month 6 to month 12 the work compounds. The pace depends on the condition of the site and the competition sitting in front of it, and the five steps above name what you are holding at the end of each stage.
Pick by who owns the thinking: a marketplace is good for one defined task, an agency is good when you want headcount on a roster, and a consultant is right when one senior person should decide the order of work and answer for it. The table above compares the three on eight things you can check before you commit.
Yes, and it is the engines reading your pages and the way the open web describes you that decide whether they do. The work is three jobs: facts an assistant can quote, access for the agents that read you, and the mentions that make you the named answer. ChatGPT is weighted first here because it is the assistant this market uses most, and the whole service sits on my AI SEO services page.
Yes, and more to the point, I get Pakistani businesses found in the markets they sell into. That means one entity described the same way everywhere, answer coverage written for each market's own questions, and the hreflang, canonical and sitemap layer that keeps your market versions reading as one business rather than several.
You own every property the figures come from, and you get impressions, clicks and conversions out of your own Google Search Console and GA4 properties, citation presence against your named prompt set, and a monthly call with the consultant who did the work. What you will not find anywhere in that report is domain rating, domain authority or brand traffic dressed up as a result.
It starts with the free 30 minute call, and what you commit to is agreed on that call and put in writing: the scope, the order of work and what counts as a result. Bring one URL you want quoted in an AI answer. If a call does not suit, send a message with the same URL and I will reply with the first finding.
Book a free 30 minute AI SEO strategy call on Google Meet and leave it holding three findings about your own site: whether the assistants are allowed to read you, which of your pages answer in their first sentence, and the first entity gap. No preparation needed beyond one URL.
Slots run Monday to Friday through the US Eastern business day, which is 18:00 to 02:00 Pakistan Standard Time while the United States is on standard time and 17:00 to 01:00 while it is on daylight time.