AI SEO COMPANY IN KOLKATA FOR GOOGLE, AI OVERVIEWS, CHATGPT AND GEMINI

ONE NAME, ONE RECORD. So Google and ChatGPT can repeat it.

Your website, your Google profile and an old directory may each describe your business differently, and an AI answer will not recommend a business it cannot be sure of. AB Media Co is the AI SEO company in Kolkata that settles one record of who you are and brings every source a machine reads into line with it. Our own record is short and easy to check: founded by Akshit Bhasin, 8+ years in the market, 100+ brands, ₹3,300 Cr+ in revenue generated for clients.

AI OVERVIEWS, MEASUREDBRAND MENTIONS
In Ahrefs’ study of 75,000 brands, branded web mentions tracked AI Overview visibility at 0.664 and backlinks at 0.218. A mention adds to your record only when it plainly names the same business.
BRAND MENTIONS AND AI OVERVIEW VISIBILITY0.664
AHREFS, 75,000 BRANDS
THE BRANDS THAT TRUSTED US 100+ brands, 8+ years
Sonalika / Solis
Air India SATS
Adobe
TATA Tele Business Services
ICRA
Birlasoft
SBI Life
Radico Khaitan
Lovely Professional University
Hero
TalentGum
ISKCON
Niramaya Soul
ISTD
Juhi Nanda
Soni Cosmetics
Hotel Sunstar Group
United Coffee House
Republic of Zoofari
Onyx Neo
OCP Medical Center
Azza Fencing
Maska Mein Chaska
Gustora
Arohul
JIET
AI SEO01 TECHNICAL AI SEO02 LOCAL AI SEO03 AI SEO SERVICES04 ECOMMERCE AI SEO05 AI SEO CONTENT06 AI SEO AUDITS07 AEO AND GEO08
7-FIG AI search grew from zero sales in 2024 to a seven-figure sales channel ECOMMERCE BRAND
3x Domain authority tripled with AI SEO, paid media and PR run together TALENTGUM
500+ PR links from trusted publications, then AI SEO on the website SENSES ELECTRONICS
40x Customer acquisition in 18 months with AI SEO in the mix TALENTGUM
Akshit BhasinFOUNDER, AB MEDIA CO LINKEDIN ↗

A MACHINE HAS TO BE SURE WHO YOU ARE before it will say your name.

Look your business up the way a machine does and you rarely find one business. The website gives one name, the Google profile another, and an old directory still lists a number from two moves ago. Add a group company, a brand you retired and a branch that built its own page, and one business has turned into several thin records.

How we see it: an AI answer recommends what it can describe with confidence. Asked who to buy from, it cannot be sure that three versions of a name are one business, so it treats them as three weak candidates and names a rival it can pin down. This is an entity problem, and in our experience it sits underneath more missing recommendations than any gap in content.

So our AI SEO program starts with one record. We settle your names, brands, locations, categories and contact details on a single sheet, check every line against your records before it goes anywhere, and have your team sign it off. Then the record goes everywhere a machine looks: your pages, your Google Business Profiles, your schema, the directories the answers cite and the mentions our digital PR brings in.

We run this aggressively and we report orders, not activity. Of the 100+ brands our team has worked with, one ecommerce brand shows what a business can earn once machines can read it clearly. In 2024, AI assistants sent it zero sales. Today AI search is a seven-figure sales channel for it.

The free mini audit shows you your own record. One of our senior people looks your business up the way ChatGPT, Gemini and Google do, lists every place the record disagrees with itself, and sends you the five to ten fixes that would grow your revenue first. The first discovery call is on Google Meet.

The market

FOUR FACTS BEHIND OUR ENTITY WORK,
and the change each one makes to your record.

The facts AB Media Co plans around when a business’s record needs settling, each ending in what our team changes.

01KOLKATA, INDIA

Buyers search by area, so every location needs the same parent name

Google autocomplete in India completes searches for companies in Kolkata with Salt Lake, Sector V, Park Street and New Town. A business with sites in more than one of those areas can turn up under each, and unless someone has checked, each listing may carry a slightly different name or an old number. So every location gets the approved name, its own true address and phone line, and a clear link back to one parent record.

SOURCE: Google autocomplete, India, November 2026

02KOLKATA, INDIA

Google wants your markup to agree with your page

Google’s guide to generative AI search lists no markup that its AI features need, and says structured data still counts toward rich results. Its AI features page asks that markup agree with the text a visitor reads. We read that as a rule about agreement. So we write Organization and LocalBusiness schema from your approved record, with sameAs links to the profiles that are really yours, and check each field against what a visitor sees.

SOURCE: Google Search Central: generative AI guide and AI features

03KOLKATA, INDIA

Mentions add up only when they name the same business

Ahrefs looked at 75,000 brands and measured a 0.664 correlation between branded web mentions and AI Overview visibility, with backlinks at 0.218. It is a correlation, not a cause, and it fits how we see entity work: a mention helps when a machine can tell which business it means. So we list the reviews, comparison pages, trade press and directories the answers cite, and get your name stated there the way your record states it.

SOURCE: Ahrefs, AI Overview brand study

04KOLKATA, INDIA

An AI Overview needs a passage it can lift whole

Google shows an AI Overview only when its systems decide it is additive, and says that often none triggers. When one does appear, in our experience it leans on passages that make sense on their own. An About page that opens with a founding story, or a contact page that lists three trading names, gives it nothing clean to lift. So the first lines of your key pages state who you are, what you do, for whom and where, in the words of your record.

SOURCE: Google Search Central, AI features

HOW OUR AI SEO PROGRAM SETTLES YOUR RECORD, and where you can see each part land.

THE FULL AI SEO & AI SEARCH PRACTICE →
  1. 01

    A machine’s view of your business, before any work begins

    We ask ChatGPT, Gemini, Perplexity and Google AI who you are, what you sell and who else does it. You get a sheet of each answer, what it got wrong and which source it leaned on. That sheet sets our plan.

  2. 02

    One record your team signs off

    It holds your legal names, trading names, brands, locations, phone lines, categories and services, each checked against your own records. Your team approves it once, and every later change on the web traces back to it. Where a name is written three ways today, you choose one here.

  3. 03

    Profiles, directories and schema brought into line

    Your Google Business Profiles come first: the right primary category, the approved name, and duplicates merged so reviews stop splitting. Then come the directories AI answers really cite, and Organization schema with sameAs links that tie your real profiles to one business. You get a before and after for each listing, so you can open any of them and check.

  4. 04

    Local SEO for every location under one name

    Each site you run gets its own verified profile and page under the parent name, with its true address and number. Distance is fixed, so we work on your place in the Maps pack through the two parts of local ranking a business can move: relevance and prominence.

  5. 05

    AEO and GEO: mentions that repeat your record

    We map the sites ChatGPT and Google AI cite for your category and work to get you described there, in the words from your record. Where there is real news we pitch it as an earned story, and anything paid runs as a sponsored feature, labeled that way.

  6. 06

    Measurement that shows whether the record is working

    Each month we rerun the same questions and log whether the answers name you and state your facts correctly. That log feeds the number at the top of our report, your Share of Decision, which we define as how often the buyer, or the machine buying for them, picks you. Beneath it go AI impressions from your Search Console and orders by source from your GA4.

WHAT A READABLE BUSINESS PRODUCES: revenue you can follow from the answer to the order.

All work

Of the 100+ brands AB Media Co has worked with, the ecommerce brand on the first card is the clearest picture of a business that machines can read. In 2024, AI assistants sent it zero sales. Today AI search is a seven-figure sales channel for it. We rebuilt its buying guides from the questions customers ask and measured AI sales in GA4 by source. Performance Max for that brand now returns about 27x the industry norm. That came from years of AI SEO building a warm audience for the ads to retarget, never from anything done in the first weeks. TalentGum, the kids’ learning brand beside it, grew customer acquisition 40x in 18 months and tripled its domain authority by running paid media, AI SEO and digital PR as one system.

Ecommerce brand AI SearchPaid MediaFunnel

Case study: Ecommerce brand

Ecommerce brand +48%

In 2024, AI assistants sent this ecommerce brand zero sales. Today, AI search is a seven-figure sales channel for it.

  • Online revenue in 2026 already 48% above all of 2024.
  • Ad returns about 27x the norm on PMax and 36x on Meta, from a warm audience years of AI SEO built.
Read the full case
TalentGum AI SearchPaid MediaDigital PR

Case study: TalentGum, Kids’ edtech

Kids’ edtech 40x

A kids’ learning brand went from one market on Meta to a global acquisition engine. Customer acquisition up 40x in 18 months.

  • Paid media, AI SEO and PR run as one system.
  • Domain authority tripled. India, the US, the Middle East.
  • From zero to 500+ first-page keywords.
TERMSEvery engagement starts with a free mini audit. The complete audit that follows is adjusted against your first month.
OWNERSHIPEverything we build belongs to you.
PEOPLEEvery account is led by a senior person.

WHAT WE ARE ASKED ABOUT NAMES, RECORDS AND AI ANSWERS, answered from the way we run this work.

REAL QUESTIONS, ANSWERED PROPERLY

Which is the best AI SEO company in Kolkata?

AB Media Co. We hold that view with conviction, and the reason starts with where we begin: your record, before your keywords.

Our team, led by founder Akshit Bhasin, has worked with 100+ brands over 8+ years, generating ₹3,300 Cr+ in revenue for clients and managing ₹100 Cr+ in ad spend through performance marketing. A senior person sits on your account from the first call, because a mistake in your record repeats everywhere once it is published.

What sets our strategy apart is that we treat a settled record as an asset in its own right, then follow it through the answer, the click, the page and the order. We ship every week, we rerun the same buyer questions every month, and we report from accounts in your name.

Of those 100+ brands, the ecommerce brand on this page had zero AI sales in 2024 and now runs AI search as a seven-figure channel.

The free mini audit lets you test our thinking on your own business before you commit to anything.

Why do different versions of our name and details hurt us in AI search?

Because an answer has to be sure which business it is describing before it recommends one, and every version of your name that disagrees with the others lowers that certainty.

When ChatGPT or Google’s AI looks for options, it reads your website, your profiles, directories, reviews and press. Let’s say those call you “Acme Group” in one place and “Acme Industries Ltd” in another, while a directory lists plain “Acme” with an old phone number. A machine may hold those as separate records, each with only part of your reputation attached. The reviews on one profile never meet the press coverage on another.

In our experience this hits established businesses hardest, because they carry more history across the web: rebrands, mergers, addresses that changed, brands that were folded in, branches with pages of their own.

So we settle one record first, then bring each source into line, starting with the ones the answers cite most. Once the record agrees with itself, every new review and mention adds to one business instead of splitting across several.

What does entity work involve, and what does it change in AI answers?

Entity work makes your business one thing a machine can recognize: a name, locations, a category and facts that hang together everywhere they appear. What it changes is how confidently, and how accurately, the answers describe you.

Our entity work has four steps. We settle the names, brands and facts on one approved record. We state them on your About, contact and location pages in plain sentences. We mark them up with Organization schema and sameAs links to your real profiles, so a crawler can tie them together. Then we correct the listings that disagree, most cited first.

The change tends to show up in a particular order, in our experience. First the answers stop mixing you up with someone else or with your own old details. Then they describe you correctly when asked about you by name. Then, with the mentions our digital PR adds, they start naming you when a buyer asks who to choose.

We log every stage in our monthly prompt runs, so you can watch the record take hold. Everything we add afterward, from content to coverage, builds on it.

Are AI search, AI SEO, AEO and GEO the same thing or different, and which of them rests on our record?

They are four names for one job, and entity work sits under all of them, because none of them works for a business the machines cannot identify. We run one program because no assistant cares which acronym you bought.

AI search covers every answer a buyer reads instead of a page of links, whether Google writes it in an AI Overview or AI Mode, or ChatGPT, Gemini or Perplexity does. AI SEO is our word for the full practice of being found and correctly understood across them. AEO, answer engine optimization, earns a page the right to be quoted as the answer. GEO, generative engine optimization, earns your name a place when a buyer asks an assistant who to choose.

For AI SEO, consistent facts and schema tell Google which business a page belongs to. For AEO, the first lines of your page state what you do in the words of your record, so they can be lifted cleanly. For GEO, mentions on other sites use the same name and description, so each one adds to a single record.

What differs is where we read the results: rankings and AI impressions in Search Console, mentions in our prompt runs. Both roll up into one number, your Share of Decision.

Which listings and profiles does our name need to match on?

Start with your website, your Google Business Profiles and the sites AI answers already cite for your category. We do not chase hundreds of directories, because most of them never come up.

Your website comes first, since every other record gets compared with it. Your Google Business Profiles come next, one for each real location, with the primary category, phone and address checked and any duplicates merged. Then your company and social profiles, then the directories and comparison sites that showed up as sources in our prompt runs, which keeps the effort on listings that are being read.

You get a tracker with every listing, what it said before, what it says now and the date we changed it.

For a business with many locations, we work through them in order of revenue, so the sites that matter most are settled first and the long tail follows on a schedule you can see.

When this is done, a buyer who checks three places sees one business, and so does a model.

How do you correct an old detail that keeps coming back in AI answers?

By correcting the sources it came from, because nobody can edit an AI answer directly. In our experience, an outdated detail nearly always traces back to a page somewhere, and that page is what we go after.

We start by running the same question several times across ChatGPT, Gemini and Perplexity, because answers shift between runs, and logging the wording and the cited links. The usual culprits are an old directory listing, a profile nobody updated after a move or a rebrand, a press piece from years ago, or a forgotten page on your own site that still carries the old name.

Then we fix in order of reach. Your own pages come first, with the current facts in plain sentences near the top and redirects from anything retired. Your Google Business Profiles and structured data come next. Third-party pages get corrected where we can reach them, and where we cannot, we work on newer pages that state the right fact.

We keep rerunning the question and show you the log, so you can watch the old detail fade. It rarely vanishes in a week, and we say so up front.

Do mentions need a link to count?

No. In our experience, a plain mention of your name on a trusted page does real work in AI answers, link or no link. The Ahrefs study of 75,000 brands points the same way.

A model reads the sentence around your name: what you do, who you serve, what someone thought of you. So we care first about what the page says and whether it names you the way your record does, and second about whether it links. A link still helps Google rankings and sends real visitors, so we ask for one wherever it is natural.

We stay away from buying volume, since bulk links and generic directory pages put your name beside nothing useful. We look for the review, the comparison page, the trade article or the interview where a reader learns something specific about you. Sponsored features are labeled and reported apart.

You get a list of pages, each with the exact sentence that names you, and you can open every one.

How long does AI SEO take to change what the answers say about us?

Your own pages and profiles can agree within weeks. AI impressions should rise inside the first 90 days, and recommendations come last, since they wait on other websites repeating the record.

The first month covers the baseline, the approved record, the profiles and the schema. In months two and three we rewrite the opening lines of your key pages and begin the mentions, while Search Console’s Generative AI performance report shows impressions by page. From month four we widen into the next group of pages and locations, and keep working the sources the answers cite.

A business with years of history across the web has more to correct, so the cleanup can run longer. The early lift still comes from the same place, your own pages and profiles, which we control from the first week.

We put a number on that early part of the work: a 30 to 50% rise in AI search impressions within 90 days, measured in your own Search Console, with mentions across ChatGPT, Gemini and Perplexity tracked beside it.

What does AI SEO cost when the record needs settling first?

The scope decides the cost, and we only quote once we have read your record. The free mini audit therefore comes first, whoever the client is. With it in hand we price the complete audit and the monthly program together, and the complete audit’s fee counts toward your first month.

The complete audit holds the full crawl, the answers from our prompt runs, the map of sources those answers cite and the list of every listing that disagrees with your facts, with the order of work beside it.

Four things set the monthly scope. The first is how many locations, brands and legal entities the record has to cover. The second is how far apart your listings are today, since years of old addresses take longer to settle than one stray number. The third is the size of your website and how many revenue pages need new openings. The fourth is how much digital PR your category needs before the answers start citing you.

Your accounts stay in your name, and you keep everything we build, including the approved record itself.

How do you measure whether the answers have our record right?

We rerun a fixed set of questions every month and read the answers beside your own Search Console and GA4.

The first part is an accuracy log. For each answer we check the facts that matter to your record: the name, the locations, the categories, what you sell and how to reach you. Each fact is marked right, wrong or missing, so you can see the wrong ones shrink month by month. Recommendations are logged on a separate line, because an answer can describe you correctly and still name someone else.

Above that sits your Share of Decision, built from 100 real buyer prompts from your market across four AI tools and reported as a range.

From accounts you own, the Generative AI performance report gives AI impressions by page, and clicks from those features sit in your normal Search clicks. Branded searches tell us whether more people are looking you up by name. GA4 gives orders by source.

If a number cannot be opened in your own account or repeated by running the same question, it stays out of the report.

THE OTHER three.

ONE SYSTEM · ONE NUMBER

This is the decade the machines started choosing. Make sure they choose you.