Iliya Valchanov

Team-GPT

Used by companies like: Maersk, Johns Hopkins University, Yale University, Shift4, EY, Charles Schwab, and many more.


Private Enterprise sofware for using LLMs.

0 to 25K users in 9 months.

Learn, collaborate, and master ChatGPT in one shared workspace.

Co-founder & CEO

Udemy

1.5M+ students in the field of statistics, data science, machine learning, and deep learning.

Bestselling courses:
- The Data Science Course: Complete Data Science Bootcamp (750,000+ students and 150,000+ reviews)
- The Business Intelligence Analyst Course (250,000+ students and 45,000 reviews)
- Statistics for Data Science and Business Analysis (220,000 + students and 45,000+ reviews)

AI Course Creator

3veta

3veta is a product company, specializing in video conferencing, scheduling, payments, and more.

Our core products are: 
→ A SaaS that simplifies scheduling and streamlines video meetings for remote and hybrid teams
→ A white-label video conferencing tool
→ API for appointment scheduling
→ Marketplace builder

Co-Founder & CEO

365 Data Science

By 2020 reached:
- 1M+ students on Udemy
- Our own learning management system
- 500,000+ students on our own platform
- 30 employees
- $1M ARR

---

It started as a simplistic WordPress website showcasing our portfolio of courses. With time, our team started growing and we developed more complex software solutions until we launched our own learning management system and developed comprehensive in-house marketing expertise.

My role was closest to that of a COO. My OCD has always been preoccupied with structuring work and developing processes in order to never miss a step or a deadline. Therefore, scaling our team and processes was more of a relief (of my obsession for structure), rather than an occupation. 

But as a co-founder I have worn many different hats with the C- in front: CFO, CTO, CMO. Nonetheless, I have never neglected the specific business driving activities of our industry, such as creating and editing content (video and text), content dissemination (automatic, semi-automatic, and unfortunately, at times, manual), social media management automation, technical SEO, tracking, ads management (running Facebook ads until my last day) and probably many more.

On the managerial side, apart from overseeing most processes, I was head of content creation and have personally read and edited all the educational content that came out of 365 Data Science. This has proved invaluable for expanding my programming knowledge base and providing me with incredible breadth of data science skills (all types of statistics and machine learning, incl. regressions, clustering, dimensionality reduction, credit risk models, price elasticity models, visualizations, feed-forward neural networks, convolutional neural networks, and more). 

In December 2020, I exited 365 Data Science to devote more time to other projects.

Co-Founder

365 Careers

365 Careers is the #1 best-selling provider of business, finance, and data science courses on Udemy. The company’s courses have been taken by more than 2,000,000 students in 210 countries. People working at world-class firms like Apple, PayPal, and Citibank have completed 365 Careers trainings.

During my time with 365 Careers, I have worked extensively on creating, managing, editing, and curating the data science content of 365 on Udemy. 

I am the author or co-author of the following courses:
-> The Data Science Courses: Complete Data Science Bootcamp
-> The Business Intelligence Analyst Course
-> Statistics for Data Science and Business Analysis
-> Deep Learning with TensorFlow 2.0
-> Python + SQL + Tableau: Integrating Python, SQL, and Tableau
-> Machine Learning 101 with scikit-learn and statsmodels
-> Customer Analytics in Python

I have overseen the creation through editing and/or project management of the following courses: 
-> The Data Analyst Course: Complete Data Analyst Bootcamp
-> The Data Visualization Course: Excel, Tableau, Python, R
-> Convolutional Neural Networks with TensorFlow in Python
-> Credit Risk Modeling in Python
-> Probability for Statistics and Data Science
-> Web Scraping and API Fundamentals in Python

I have also lead the translation effort of video content in German, Portuguese, Spanish, and Chinese, resulting in the following courses:
German:
-> Das komplette Data Science Bootcamp 2021 in Python
-> Statistik für Data Science und Business Analytics
Spanish:
-> Bootcamp Completo em Data Science com Python 2021
-> Estadística para Data Science y análisis de negocios
Português:
-> Entrenamiento completo en Data Science con Python 2021
-> Estatística para Data Science e Análise de Negócios
Chinese:
-> 数据科学和商业分析中的统计学

Content Manager

CEO @ Team-GPT | 1,300,000+ AI students on Udemy

👉1 sentence about me: 
I teach AI to 1.5M students on Udemy; also CEO of Team-GPT: A leading Enterprise AI software. 

👉Multi sentence about me: 
I used to compete in Maths for 10 years and was quite good at it. To this day, my work ethic is highly influenced by my Math competitor days.

I graduated in Finance and have all the financial knowledge I could ask for. It was not my calling, so I never pursued it... but I find it quite useful in running a business. 

At 23, I co-founded an educational startup for online data science courses - 365 Data Science. I helped it scale to 30+ people with millions of students worldwide and $1M+ ARR. Courses that I have personally created have been taken by 1,300,000+ students. This EdTech business was bootstrapped to $30M+.

Currently, I'm Co-founder and CEO of Team-GPT. Team-GPT reached 25K users in 9 months and is used by enterprises all around the globe.

I'm a very curious guy, who loves to build stuff.

Learn, collaborate, and master ChatGPT in one shared workspace.

You got into marketing because you love the work. 

The strategy. The creative. 
The moment a campaign clicks and everything comes together.

But that’s buried. 
Dozens of tools. Hundreds of workflows. Endless details.

You spend more time managing systems than creating. 
More time switching contexts than turning ideas to life. 
More time explaining what you're doing than actually doing it.

This isn't what you signed up for.

Marketing should be simpler. 
Much more delightful.
More like... Juma!

-

Juma is the workspace where your team does their best work.

Everything you need to plan, create, and launch lives in one place. 
Your team has the same context. The same tools. 
The same ability to move fast. 

A place where work stops feeling like a burden.
Experience the joy of marketing.

Think together, move as one!
Welcome to Juma.

juma.ai

CEO @ Juma | 1,500,000+ AI students on Udemy

Когато пиша за политика, алгоритъмът се обърква, тролове коментират, маркетинг стратегията ни пропада, а бизнесът ни страда.
Но има нещо от което бизнесът ни, както и на много други българи, страда повече.

Това, че ни спират тока, като им пречим.
Това, че съдебната система е завладяна от мафията.
Това, че взимат бизнеси на хора безнаказано (Илия Златанов от Изамет и Осемте Джуджета, както и десетки неразказани истории).
Това, че държат кмета на Варна 6 месеца в затвора без причина и без присъда.
Това, че саботират полезните действия на кмета на София.
Това, че Европейската прокуратура трябва да се намеси, за да търси справедливост срещу кметове на ГЕРБ и безумията им. 
Това, че САЩ трябва да използва Магнитски, защото сами не можем да се справим с корупцията.
Това, че от години в България няма истинска журналистика, защото разкараха всички достойни журналисти.
Това, че медиите показват каквото им се каже, а социалните мрежи показват истината.

Това, че закърпват държавния бюджет с парите на най-здраво работещите хора. Не за да ги дадат на учители и лекари, а за да приберат парите в джобовете на безкрайно оядени престъпници. 
 

Ние сме хората, които градим, а не рушим.
Които вкарваме външни инвестиции в страната и увеличаваме богатството на България.

Протестираме, защото заслужаваме много повече от прасе и тиква.

devote-academy

LinkedIn

li_at

JSESSIONID

Talks about #ai, #chatgpt, #startup, #buildinpublic, and #entrepreneurship

New product, who dis?

Soon to empower every marketer in New York.


Proudly created this in Juma

Now 15,000+ people follow my content here.

I'm the CEO of an AI startup but don't use AI to write on LinkedIn.

Why?
-> I think AI can help with wording.
But it can't help with self-expression.

-> Lots of work will be automated by AI.
But human connection is not one of them. 

-> AI writes cheaper, faster, better than me.
But I mainly write to structure my own thoughts.


That's why I'm on LinkedIn -> 
to structure my own thoughts, express myself, and hope to connect with likeminded people on the way. 


In the last 3-4 years of posting, I shared the good, the bad, and the ugly.
-> Reached millions of AI students with 365 Data Science 
-> Closed 3veta and shared why
-> Built Team-GPT in public from Day 1: 0 to 10K users posts, fundraised $4.5M, posted investor reports, shared deeply dramatic insights, hit $1M ARR, and even posted the OpenAI legal letter suing us


I share because 𝘺𝘰𝘶 have shown me it's useful!
And am thankful for your time. 

.
.
.

P.S.
If you don't know me -> let's connect!
If you know me -> Thanks for being around, I'll continue sharing 🫡

This weekend I finally did it. 
I've been avoiding it for 2 years.
I'm planning for it to last 50+ years. 

It's my personal website, featuring:
- links to my work
- some of my favorite articles and posts
- mini projects (e.g. beautiful QR code generator)

I have a bunch of stuff I've been meaning to publish online but LinkedIn doesn't feel like the right place. 

Check it out 👇

iliya (.) ai

graphical user interface, application, Teams

Първи ден на Джума и вече е в новините 😎 

Благодаря на Mary Ivanova и Капитал, че отново разказаха историята ни!

🤖 Преди година, малко след като #TeamGPT привлече 4.5 млн. долара финансиране в рекордна за млада компания на този етап сделка, в лондонския офис на дружеството пристига писмо, адресирано от OpenAI. "В него ни подтикваха да се откажем от търговската си марка в САЩ, Великобритания и ЕС, защото не можем да я притежаваме", обяснява Iliya Valchanov, CEO и съосновател на Team-GPT (българското дружество се казва "Тийм ГПТ").

📧 Екипът от петима бизнес партньори решава да игнорира писмото, защото компанията има запазени марки на трите пазара, одобрени от съответните контролни органи. Разбира обаче, че не можеш да игнорираш OpenAI така, "както можеш да не отговориш на съобщение на мъж след неуспешна среща", смее се Вълчанов...

🌍 Тази седмица стартъпът, който от две години обслужва вече над 75 хил. клиенти в САЩ, Канада и Европа, сменя своето име и търговска марка. Team-GPT вече се казва Juma. Какво се случи междувременно, какво правиш, когато технологичен гигант съобщи, че иска да те съди и какво казва законът? Четете в статията  на Mary Ivanova тук: https://lnkd.in/dJ-2qRqy

Open article: Team-GPT вече е Juma: какво правиш, когато OpenAI иска да те осъди by capital.bg

Team-GPT вече е Juma: какво правиш, когато OpenAI иска да те осъди

Founders, VCs, angels:
Our fundraising materials for raising $4.5M in San Francisco as Silicon Valley outsiders.

What it took? 
85 'NO's and a physical breakdown.

The steps?

Prep:
1. Read books
2. Meet founders
3. Prep materials
4. Get angels
5. Get VCs

The VC bits:
1. Create an investor list
2. Get warm intros
3. Pitch deck is used to get a 1st meeting
4. NEVER show the pitch deck in the first meeting
5. 1st meeting is to get a 2nd meeting
6. Follow-up with the f*cking best materials
7. Receive a term sheet or get a ‘NO’ ASAP


Unfiltered, 1-hour lecture:
👉 https://lnkd.in/dt6xEJhB

Other materials:
1. PPT presentation from video: https://lnkd.in/dhpirdmD

2. How to raise $5M-$15M from Silicon Valley? Learnings from a conversation with a VC from Sequoia Capital: https://lnkd.in/ggPfyJFh

3. How Team-GPT raised $4.5M- A True story - 100% real story about our fundraising. Journalled as it was happening.
https://lnkd.in/dZ3UxmfE


Thanks for the support along the way!

Our product is largely shaped by a couple of visionary clients.
VertoDigital is one of them.

Amazing webinar with their team starts in 1 hours. Don't miss it!

👉 https://lnkd.in/dmbh2QP3

No alternative text description for this image

Getting from 0 to 10K users was wild.

Revenue: $80,000
Net revenue: $ 47,819.58
Time: 170 days
Users: 9597

Who remembers the first days of building Juma (ex Team-GPT) in public?

- Took us 6 months to make $80K in total revenue.
- Last month we added $80K recurring revenue. 

The scale has changed drastically. 
But we still have the same passion! 
And much better direction.

Here's how we got from 0 to 10K users:
👉 https://lnkd.in/gFXPFhga

Yellow alert!

We are changing everything about us next week.

New name, new agents.

Stay tuned!

Not all $$$ are made equal. 

Recurring revenue is way more valuable than one time. 

Service revenue is easier than product revenue. 

B2B revenue is easier than B2C. 

Enterprise revenue is hard to get, hard to lose.


The goal of each business?
It's not generating money but generating free cash flow.

So it's not about how much money you make. 
It's about how much money is left.

86% of marketers believe AI will benefit their work.
Only 26% can actually execute on it.

That's a 3x gap between confidence and capability.

Many teams are stuck:
- Not decisive enough to commit. 
- No internal drive to manage the adoption.
- Not technical enough to set AI up for success.


The Growth Syndicate made a report which covers 110 B2B marketing leaders and breaks down:

1. Where teams are actually using AI (91% for content, only 20% for sales enablement)

2. Why skills matter more than budget (60% cite training needs vs 25% budget constraints)

3. Much more


My contribution to the report:
"Don't be afraid to make changes to your strategy as you learn more. But with any marketing activity, if you give up after one post on LinkedIn, have you really tested the channel? 
Same with AI tools—commit long enough to learn, then course-correct."

Link:
https://lnkd.in/d3seAUE7

To make $1M per year,
you need:


1,000,000 customers paying you $1
100,000 customers paying you $10
10,000 customers paying you $100
1,000 customers paying you $1,000
100 customers paying you $10,000
10 customers paying you $100,000
1 customers paying you $1,000,000


Which one do you prefer?

Next week AIEF After Hours is in our office. So I couldn't miss it, even if I wanted to.
But I wouldn't miss it because Rangel Milushev (YC-founder, Harvard gradudate) will be talking about AI.

You should totally come: 
👉

Open article: AIE.F Europe After Hours #46 · Luma by luma.com

San Francisco hits different when you're all-in building.

Sure, the investment money helps. 
But it's not about that.

It's about the energy. 
The ambition. 
The history.

The belief that you can actually change the world with a laptop and a vision.

You're the average of the 5 people closest to you. 
When that may be Elon Musk, Mark Zuckerberg, Sam Altman, Satya Nadela, and Jensen Huang you either become a supervillain or you get really ambitious. 

Hope it's the latter.

We're cooking up something that will change how people market stuff.

More soon!

I know these tease posts have been a bit too much...

But I literally have only 2-3 days left to do them.

Stay tuned- 
dropping our biggest update ever next week!

How do you know your outbound sucks?

It's outdated.

I still receive calls for 365 Data Science. 
I receive emails for 3veta.
LinkedIn messages for Team-GPT.

Idk where these people get their data but a message like this kills trust immediately. 

0% chance to work together.

graphical user interface, text, application, email

'Dogfooding' means being our own customer.
All successful startups practice dogfooding.

I'm in top 1% of the users of Juma, using AI more than 99% of the people on the platform.

Juma (ex Team-GPT) has saved me ~2 months of work in 2025. 

My favorite capability? 
Projects!

I hate Black Friday.

After building multiple products and runnning many promotions-
Black Friday deals work only for ecommerce.  

- For software, it cheapens the product and the experience. 

- If you do this often, people get used to it and 'wait for the discount'. 

- Loyal existing users feel betrayed that you gave a deal to someone else, while that someone else only chose you cause you are cheap, not because of your value prop. 

+ the revenue is not that much anyways

There is a lot of hype in AI because there has been a big shift in what's possible. 

It's clear a lot of value is being unlocked. 

Truth is- we are all figuring it out. 

👉 We, the builders, are trying to build new age tools which create 10x or 100x more value than existing software for our users. While figuring it out, we are often creating more noise and not making it easier for users. 
(when you have more clarity a rebrand often comes, see us as an example haha)

👉 For AI adoption leaders it's absolutely clear they need to act. But what to do? The best results are seen when:

1) There is a clear owner who is responsible for making sure people in the org use AI. This person should be empowered to make decisions, rollout the way they see fit, and keep going with this decision for at least 3-6 months. 

2) Don't be afraid to make changes to your strategy as your learn more. But similar to (say) LinkedIn posting, if you give up after you post 3 posts on LinkedIn, have you really tried?
Same with the AI tools- 𝗰𝗼𝗿𝗿𝗲𝗰𝘁 𝗰𝗼𝘂𝗿𝘀𝗲 𝗯𝘂𝘁 𝗱𝗼 𝗰𝗼𝗺𝗺𝗶𝘁.

3) Do weekly check-ins with people using these tools. The fact that a tool has been purchased, doesn't mean it is being used. Be on top of it and measure. Make data-driven decisions. 


AI is not overhype.
NVIDIA, OpenAI, and similar are. 

But AI is creating a lot of value. 
But it needs to be harnessed.

GTM for Early Stage SaaS | Co-founder @ Datapad | SaaS Growth Advisor | Reforge | 1x Exit |

I’m a marketer, 2x entrepreneur with 1x exit, and helped companies to go from 0 to 8 figures.

My achievements:

- Built 2 VC backed startups with one exit

- Built a 7 figure productized marketing agency

- Became a mentor in top accelerators such as: 500, Plug and Play, Techstars

- Got to work with the portfolio companies of amazing investors such as A16z, Craft Ventures, YC

- Built a podcast with 20k downloads in a year-figure

- Transformed my body more times than Christian Bale ☺️

John Ozuysal

Hiring Alert: I am looking for a GTM engineer (contactor)

This is not a client-facing role; you will not be running campaigns for clients, but you will be helping me run internal fun experiments. No middleman, just 1-1 with me.

You have to be good at: Clay, Smartlead, Instantly, Trigify, etc...

Your main responsibility will be to run internal campaigns with me to test different offers

Most importantly, you should be able to catch up with my tempo :) 

I will be in Dubai for the next couple of months, and ideally, you should be in a similar time zone.

If you want to learn from someone who has built and exited SaaS companies and scaled agencies to 7 figures, DM me and I will send you more info.

Generate Pipeline from AI Search | 1x Exit

I would rather start LinkedIn commenting as a service, rather than content writing.

I am not sure if it’s just me but comments are getting much better engagement then posts.

Even my mum is not seeing my posts anymore 😀

Every time I open LinkedIn, I see less expertise and more theatrics. 

And now it is all boosted around the clock.

-People “proving” AI search visibility by posting… their Google Search Console
 
- Others claiming they “ranked in AI in 14 days,” but all they show is a Google SERP and a couple of AI overview citations

 - Tombstone hobbyists declaring a new marketing channel dead every single week

 - Creators making cringy YouTube skits hoping it magically turns them into experts

 - Founders flexing “10M ARR in 5 months” — in what currency, nobody knows

 -Endless fake social proof screenshots that fall apart the moment you zoom in

 And now… boosted posts from these same people flooding our feeds 24 hours a day

For anyone who understands basic marketing, LinkedIn has become a loop of fake promises, fake authority, and zero practical value.

The platform used to be where real operators shared what actually works.

Now it's mostly noise amplified with ad spend.

Is it just me feeling like this ?

Here is the 4-week plan I would follow to get AEO visibility in 4 weeks

👇

I can't say that I cracked the code, I think no one can but below is the plan I am using the get results fast

👉 Grab the full play here for FREE: https://tally.so/r/jaDPl9

WEEK 1: IDENTIFY LOW-HANGING FRUITS

First, we identified the high-intent prompts that we should go after. We found several Reddit threads being cited. We created content on these threads and pushed our comments up for more visibility.

We initially started with 4 Reddit threads.

WEEK 2:UPDATE EXISTING PAGES

We updated some of the BOFU pages to make sure that the content is optimized for Google and AI search engines.

Pages were terribly structured, so we did rewrites first ( if the team is reading, sorry, but you know I mentioned this before we started to work )

After the rewrite, our updates were focused on:

- Adding data and statistics

- Query fan-out optimization 

- Entity optimization under every H2 

- Adding fresh information (recency is very important for AI)

- Tables for product comparison

WEEK 3: 5 HIGH INTENT PAGES

2 alternative pages against their top competitors ( we like to start with vanilla alternatives content, and we verticalize the alternatives articles by creating them for each persona)

2 three-way comparison articles

We also did a competitor pricing review article ( this one actually landed us a demo with a big enterprise co. They did not book a demo, but we identified the visitor and reached out )

WEEK 4: OFF-SITE MENTIONS

- We worked on repurposing our 2 pieces of BOFU content and uploaded them to YouTube. 

- Secured 3 mentions where LLMs were citing the source

- Added 4 more Reddit mentions - we are using Scrunch to keep track of threads that are influencing the prompts we are going after

-----

👉 Grab the full play here for FREE: https://tally.so/r/jaDPl9

I am looking for a GTM engineer a part-time or project basis

This is not a client-facing role; you will not be running campaigns for clients, but you will be helping me run internal fun experiments

You have to be good at: Clay, Smartlead, Instantly, Trigify, etc...

Most importantly, you should be able to catch up with my tempo :)

Listicles help with getting AI search visibility or no?

Here is the research 👇

Ahrefs have just done and published a research piece on AEO that all of us knew but couldn't prove:

Self-promotional listicles (best X software/agencies) are one of the best ways, if not the best way, to be the #1 recommendation spot for a solution comparison.

According to Glen Allsopp's research that analyzed 26,283 source URLs on AI search:- “Best X” blog lists represented 43.8% of all page types.

- Placing higher on multiple comparison lists likely increases your chances of being recommended by ChatGPT (which is why I've started doing outreach to get mentions).

- Across 1,100 blog lists used as research sources, 79.1% were updated in 2025 (which only confirms how important content freshness is).AI search for SaaS brands and agencies, put in the most simple way possible, consists of 3 core pillars:- Owned content, which covers comparison guides, where you explain what makes you the #1 option in the industry or for a specific use case.

- Mentions, which covers having visibility on other comparison lists, review platforms (e.g., Clutch and G2), and community forums like Reddit.

- Content freshness, which I think is AI search engine's ''sanity check'' if your content is outdated or recent.Here are a few examples of owned content:

- Top 10 [Competitor] Alternatives For ITSM In 2026.

- Top 10 Project Management Software For Start-Ups In 2026.

- 10 Best Marketing Agencies In Dubai For SaaS Brands In 2026.

And then positioning your solution or service as the #1 choice, clearly explaining why, and then covering other competitors in the space.

Glen's research also found something I have been thinking about for some time but could not prove:''35% of cited best lists were published on low authority domains.''Start-ups and low domain authority domains CAN WIN in AI search, and should not be scared of competing against established players on the market.

This research confirms the trends we have been seeing with the portfolio companies of House of Growth - AEO Agency

Link to research under comments

We helped Juma close 6-figure deals from AI search

Get the playbook for free (no need to comment) 👇

This is the same playbook we used to create a pipeline for all of our clients. You can take it and implement it in your business straight away.

Our AEO playbook includes:

- Everything we know about AI models' recommendation algorithms

- Differences between Google and AI search behavior 

- How to ensure models can properly access and read your content

- Best practices to get recommended on AI models (including brand mentions, BOFU content creation, and more)

- Real examples of content that works (and what fails)

👉Grab the playbook from here:https://tally.so/r/jaDPl9

Reading too many SaaS trends and predictions by people who have built or worked inside a SaaS company 😅 

I don't understand why people try to fake authority to gain likes.

Don't listen to everything you see here. Always rely on your own playbooks and keep testing!

Stop letting your competitors control the narratives inside ChatGPT

👇

ChatGPT isn’t "hallucinating" about your SaaS. It’s improvising because your website is a ghost town. You are not training AI with your own story.

Here’s the harsh reality:

LLMs are prediction engines. When you leave gaps in your digital footprint, the models fill them with whatever they can find: competitor comparison pages, angry Reddit threads, or 3-year-old press releases.

If you want to control the narrative in the age of AEO (Answer Engine Optimization), you need to spoon-feed the models.

You need to move beyond marketing fluff and build a "Source of Truth" architecture.

Here are the 3 types of pages I would build today:

- About us: Make sure that it's very clear on what your company is doing, who is your ideal ICP etc. Avoid using fancy language to impress your investors.

- Comparison pages: Control the narrative by owning the comparison-related queries. If you don't do this, someone else will do it, and AI will spit out the wrong information about your company.

- Integrations page: People don't just search "best project management software anymore; they search " best project management software with integrations to x". It's time to surface your integration docs and make them visible to search

If you don't provide the language, the model will generate its own. And I guarantee you won't like the version it writes.

Warmly, ($5m +ARR ) generates ~25% of its pipeline from AI search.

We have been running their SEO / AEO, and here is the playbook 👇


1. What we've been doing so far: Large-scale BOFU content creation:

- Competitor alternatives content: [Competitor] Alternatives
- Category keywords: Best [Category] Software
- 3-way comparisons: [Competitor] vs. [Competitor] vs. Warmly
- Pricing guides: [Competitor] Pricing (These guides work wonders for competitors that do not disclose their pricing on their website).

Covering the same competitors in these 4 formats allows us to build what I call competitor topical authority to help us rank better for competitor-related keywords.

Plus 

I am still a big fan of focusing on "information gain score," so we always aim to create content with more relevant information than competing pages.

2. What we've been focusing on in Q3 2025: 

Vertical comparisons + Reputation management/building + Adding schema markup + Updating BOFU pages that rank lower than position #3 on Google

VERTICAL COMPARISONS

Focusing on creating highly-specific BOFU content for certain ICPs (e.g., marketing agencies).

Example: [Competitor] Alternatives for Marketing Agencies. 

I know this does not have a reasonable search demand to go after it. But our big bet here is that users are going to provide context about their solutions when evaluating alternatives and solutions on the market.

Especially when using ChatGPT's ''Deep Research'' I've seen from user interviews (yes, I actually ask people we work with to share their prompts) that they share information about their industry or profession so they can get a more tailored answer.

Hence, why we're going to give them the answer to what is the best competitor alternative is for each persona we go after

REPUTATION MANAGEMENT & BUILDING

With OpenAI making deals with Reddit and mentioning Quora or other media publishers in its sources section, we cannot sit idle and expect people to happily discuss us.

This is why we'll be doubling down on these 4 channels

- Reddit.
- Quora.
- YouTube.
- Media publishers.

The goal will be to get mentioned there and explain what makes us stand out from our competitors.

SCHEMA MARKUP

We're planning on implementing schema markup on our website on our existing and future content to better explain our organization and product to both Google and AI search engines.

CONTENT FRESHNESS

After a bit of experimenting, I've noticed that AI search engines really like to favor fresh, up-to-date content that's been updated recently.

The same goes for Google nowadays as well. The more we update our content and keep up with the latest information (e.g., updated pricing, new features, new competitors), the more likely crawlers are to pick up our content and recommend it.

Generate Pipeline from AI Search | Founder @ House of Growth | 1x Exit

Here is why SaaS companies losing clients to competitors on AI search

Your competitor just taught ChatGPT what to say about you.

And you weren't even in the room.

Here's what happened:

Last week, a prospect searched "best email marketing tools" in ChatGPT. Your product didn't show up. Your competitor's did.

Not because their product is better.

Because they own the pages about their category and competitor review content. You didn't.

I see this constantly with early stage B2B SaaS. They spend six figures on ads and PR, then wonder why AI is recommending their competitors.

The answer is brutal: you left the door open.

LLMs are scraping everything right now. Every product page. Every comparison chart. Every "X vs Y" article on your site (or your competitor's site).

If you're not publishing this content, you're not in the conversation.

Here's what you should do:

Write a "best [category] tools" article

Create comparison pages for every competitor. 

Don't forget to add three-way comparisons, too

Done? Go and turn these into videos and publish on YT as well.

Because if you don't explain what you do and how you are different than competitors, someone else will write that and control the narrative.

It's about being present when the buying decision happens.

My clients start seeing traffic from AI sources within weeks. Not months. Weeks.

The best part? Your competitors probably aren't doing this yet. They're still waiting for someone to write about them.

Control your narrative, or watch someone else write it for you

I watched a $50M company disappear from AI search in 90 days.

Not because they got penalized. Not because their SEO team quit.

They just hit pause.

Three months later, their biggest competitor owned every ChatGPT answer in their category.

The company that paused? They're still trying to catch up.

And here's the part that keeps me up at night:

They thought pausing was the safe move. Budget cuts. Uncertain ROI. Let's wait and see.

But in AI search, there is no neutral.

Every day you're not feeding the machine, your competitor is.

Every article they publish trains ChatGPT to recognize THEM as the authority.

Every brand mention compounds their lead.

Every Reddit thread you miss is an opportunity for a competitor

And the gap doesn't just grow. It accelerates.

Because freshness wins now. And whoever shows up consistently becomes the default answer.

Your buyers only used Google to start their searches, not that they are switching to LLMs

And if you're not in that conversation, you don't exist.

Standing still isn't playing it safe. It's surrender.

The companies winning right now aren't the ones with the biggest budgets.

They're the ones who understood 6 months ago that AI search isn't coming.

It's already here.

Your competitors are guessing AI search prompts. You could just ask.

Here is how 👇

No tool is going to hand you the exact prompts your customers are typing into search.

But you can figure it out without paying for a single piece of software.

Looking back, my best keyword ideas did not come from keyword research tools. 

They came from customer interviews and simply asking, “How did you find us?”

Right now, everyone is trying to hack their way into uncovering the most valuable LLM search prompts.

Go back to the fundamentals.

Here is what I tell my own companies:

 1️⃣ Ask every new customer if they found you through an AI search engine.
 2️⃣ If yes, ask exactly how their search started.
 3️⃣ Repeat this for your last 100 prospects.

You will start spotting patterns and those patterns are your high intent prompts.

Once you have that list, track your visibility for those prompts and work to win those spots.

This is how I am doing it. You should try it before your competitors figure it

taplio

The Battle

Iliya Valchanov

John Ozuysal

Followers

Posts

Engagements

Likes

Performance ScoreIliya Valchanov's LinkedIn Performance

Performance ScoreJohn Ozuysal's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Iliya Valchanov

John Ozuysal

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Iliya Valchanov's LinkedIn Performance

Performance ScoreJohn Ozuysal's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreIliya Valchanov's LinkedIn Performance