Skip to main content

AlgoEgo

sudhir

AI search and customer trust advisory by Sudhir Sharma

I’m Sudhir Sharma. Google hired me as an SEO in 2016, the search company hiring for search, and I spent the next decade inside. Today I advise on AI search and customer experience, with 17 years in search behind me. I help leadership teams with the two judges every company now answers to: the algorithms that decide whether you’re named, and the people who decide whether they trust you.

Before Google, I built Adobe’s SEO consulting practice for Marketing Cloud customers, from startups to the Fortune 100, and led SEO for BrightEdge. At Google I worked across customer experience, product and growth in APAC, including crisis management for YouTube, where almost every grievance was about the algorithm.

I’m the founder of Hupik, an AEO agency, an adjunct professor of digital marketing at SUSS, and hold an Executive Master in Change from INSEAD.

My focus is AEO, GEO and AIO: answer engine optimisation, generative engine optimisation, and visibility in Google’s AI Overviews. Alongside that, E-E-A-T and entity clarity, how AI engines represent brands, customer grievance and crisis, and the human side of putting AI into customer service.

17yrs

In search

10yrs

At Google

3

Employers acquired

4

Countries with live work

Experience across

The thesis

Every company now has two judges.

The Algo decides whether you get named. The Ego decides whether you get chosen and, when something goes wrong, whether they felt heard. Optimise for the Algo alone and you’re forgettable. For the Ego alone and you’re unfindable.

Work with me

Four ways to work with me, depending on what you need next

01ALGO

Get Found

Find out what ChatGPT, Gemini and Google say about your company, and fix what’s wrong. I run the searches first. Then I tell you the three things worth doing.

For: Leadership and marketing teams

02ALGO

Learn It

Want your team to understand AI search, not just pay for it? I teach it in plain English, from a decade inside Google and my teaching at SUSS.

For: Teams who want to do it themselves

03ego

When It Goes Wrong

Complaints, bad reviews, a crisis, or AI about to go in front of your customers. I help you fix the human side first. It’s what people remember, and what the algorithm reads.

For: Leaders responsible for customer trust

04both

Work With Me

Some problems sit across search, customers and change, and don’t fit one box. Tell me what’s happening. I’ll tell you straight where I can help, and where I can’t.

For: Problems that don’t fit one box

Questions executives ask me

How is AEO different from SEO?

SEO earns you a place in a list. AEO earns you a place in the answer. Google’s results page ranks pages and lets the customer compare; an answer engine does the comparing for them and names a shortlist, sometimes one company.

The inputs overlap: a sound site, clear content, consistent facts. The weighting shifts. Rankings reward your pages; answers reward what the whole web says about you: press, analysts, reviews, Wikipedia, forums and structured data. You can be top of Google and absent from the answer. That makes it a leadership problem, not a ranking problem: the answer about your company is written by a system nobody inside it controls.

Four names, one question

SEO · search engine optimisation: a place in Google’s ranked results.

AEO · answer engine optimisation: being the answer an engine gives.

GEO · generative engine optimisation: cited and described accurately when a language model writes a response.

AIO · usually Google’s AI Overviews, the summary above the results; some vendors use it for “AI optimisation”.

Evidence before opinions. I start with a prompt panel: the 20 to 30 questions your buyers ask before they know your name, run across ChatGPT, Gemini, Google’s AI Overviews and AI Mode, and Perplexity, and repeated, because the answers change. I record who’s named, in what order, what’s said, and which sources each engine cites.

That shows three things: whether you’re in the answer, whether what’s said is true, and which sources are shaping it. The work then goes where the engines actually read: pages that answer the question in their first lines, entity and structured data so the engines know exactly who you are, and the third-party sources they trust.

Then I pick the three things worth doing, not the thirty that could be done.

Against the baseline, on the same panel, every [month]:

Share of answers

How often you’re named across the panel, engine by engine.

Position

Named first, in the shortlist, or a footnote.

Accuracy

Whether what’s said about you is true and current.

Citations

Which sources the engines quote, and how many of them are yours.

Business signal

Referral traffic from AI engines in analytics, branded search, and enquiries that say an AI sent them.

Rankings and traffic still get reported. They’re no longer the scoreboard.

The AI representation briefing is this method, run on your company. Book the briefing →

What do your angry customers teach the algorithm?

An unresolved complaint doesn’t stay in the queue. It becomes a review, a forum thread, a post, and then a line in an AI answer about your company. At YouTube, almost every grievance was about the algorithm. Now the algorithm is reading the grievances.

That makes reputation management part of AEO, not a separate PR job. Your reviews are what the engines read. And trust isn’t built by the complaints process; it’s built by whether the person feels heard before they ever need one.

AI doesn’t remove the anger from a process that already fails people. It industrialises it.

Book a consult

Which judge is the problem?

If you already know, book a consult. If you don’t, tell me what’s happening and I’ll help you work out which.

I reply personally within two working days.

    Not ready to book? Write to me