On Creating An Ad Management Platform Serving 100B Monthly Ad Requests

Published: March 24th, 2020
Rajiv Khaneja
Founder, AdButler
from Victoria, British Columbia, Canada
Discover what tools Rajiv recommends to grow your business!
social media
Discover what books Rajiv recommends to grow your business!
Want more updates on AdButler? Check out these stories:

Hello, my name is Rajiv Khaneja, and I am the founder of multiple SaaS companies, including AdButler.com. I am also an angel and a real estate investor.

AdButler helps the web's most creative publishers monetize and measure their content. We currently serve over 100 Billion ad requests per month for over 1,500 publishers.


What's your backstory and how did you get into entrepreneurship?

Like a lot of entrepreneurs, growing up I always had money-making schemes, renting out my video games, weeding gardens, paper routes, building computers, selling hockey cards, etc.

Be wary of mediocre success. Mediocre success is way worse than failure. While failure will sting for a while mediocre success can torture you for years or decades.

In grade 4 I took my paper route money and began trading stocks and I learned how the business worked.

Eventually, that entrepreneurial desire combined with my computer hobby and I stumbled into entrepreneurship.

Take us through your entrepreneurial journey. How did you go from day 1 to today?

In 1995 I was 13 years old, growing up in Victoria, Canada, far from Silicon Valley at the dawn of the dot-com boom. I started playing around with the Internet and was blown away. The idea that I could create a website, upload it somewhere and someone half a world away could see it struck me as insanely cool.

Fairly quickly, I learned how to make these magical "homepages" and became obsessed with getting hits. I made a dating site, a jokes site and a bunch of other stuff.

Honestly, I was just having fun. I had a secret life that nobody understood. My parents, teachers, and friends didn't have the Internet, know what it was, or had any idea what I was up to. On a typical day, after going to school and playing hockey with my friends, I would come home and delve into my secret world.

Reasonably soon, my websites started to catch on, generating hundreds, then thousands of "hits." Each time I'd check the site and see the hit counter jump up, I'd get a little dopamine hit. It was like a drug, and I wanted more.

Right around that time, the first internet advertising started to emerge and I realized that I could actually make money doing this. I remember finding my first advertiser, the Pacific Coast Down Company. I think they paid five cents a click.

I sat back and did the math. "If only 10% of the people that come to my site click on this banner, I'll make hundreds of dollars per month". I told all my friends, and they all thought I was insane.

Remember this was a very different time. Making money on the Internet was not a thing. The idea of putting your credit card into the computer was insanity. Online dating was for deviants and creeps.

Back then, there were no analytics, no dashboards or any way to see your advertising performance. The Pacific Coast Down Company-paid "net 90". That meant I had to wait 90 days to find out how much money I made.

I have a very vivid memory of my Dad holding a white, already opened envelope. "This came for you," he said, handing it over. I was too excited to care that he had opened my mail. I grabbed the envelope and pulled out the check... and just stared at it. The whole check was for... $6.03

I didn't give up, though. I quickly realized that if X visitors to my website resulted in $6, then 10x would be $60, and 100x would be $600. And that was a meaningful amount of money to a kid who's the sole source of income was a $10/week allowance.

I was constantly tweaking the site, posting everywhere I could to get the visitor count up. I realized I needed a reason for visitors to return. I started a "Joke a Day" section and a daily joke email list.

One idea I had was to put a Poll on my site. I figured I could ask provocative questions, and those who voted would then want to return to check the results and see new questions.

I knew HTML and figure adding a poll would be easy. There must just be a <poll> tag. But after a quick AltaVista search, I realized it wasn't that simple.

There were thousands of people asking the same question, "How do I put a poll on my website," and we were all getting the same answer "You need to learn to program" ... "You need to learn something called Perl."

Ok, I thought Perl is probably something like HTML. So I looked up Perl. It wasn't like HTML at all. While HTML was readable and understandable, Perl was complete gibberish. I was confronted with an incomprehensible stream of dollar signs and exclamation points and other nonsense. So I gave up on that crazy idea.

But the idea stuck with me. I really wanted a poll on my website. I knew it would be a hit, and I really wanted to make more than $6.

So one weekend, I locked myself in my bedroom, determined to figure out Perl.

I tackled the problem sideways and took as many shortcuts as I could. I didn't actually want to learn Perl at all, I just wanted a Poll on my website.

It occurred to me that a hit counter, something that increments a number by one each time it is loaded, was pretty similar to a Poll. I just needed multiple hit counters, one for each option someone could vote on. And the counters needed to increment conditionally based on a user's vote.

So I downloaded a basic hit counter script and hacked it. I opened the file that makes it run and began experimenting. I'd change something, upload the file and run it. 97% of the time, I'd get some horrible error message (well actually the same horrible error message there was only one "500 internal server error") and quickly hit undo. 2.5% of the time, it would run but do something entirely unexpected, and 0.5% it would do what I wanted.

Through this learning to code by trial and error process, I eventually created what was probably the world's worst poll script. But the quality didn't matter, what mattered was that it worked, and the Polls were a hit. Generating tens of thousands of votes.

Shortly after that, I had a realization, I remembered that there were thousands of other people looking to put a Poll on their site, and they were all getting the same answer "You have to learn Perl."

I figured that if I just made a small change to my script, I could allow other people to create Polls on their websites for free, and in exchange, I could put a banner ad on their poll results page.

I did that, and it absolutely exploded. Everybody wanted a Poll for their website.

It was lit by fire. My website quickly crashed, my hosting company kicked me off, and I was suddenly forced to learn things like algorithmic efficiency and SQL databases. Then, fueled by dot-com 1.0 VC cash, banner ads started paying off. Big time.

Before graduating from high school, I made enough money to buy a house and retire.

I had learned from my immigrant parents the value of living below your means and saving money. So that's what I did. I was also taught not to flaunt success and never to show off. So I did that too.

At one point, a reporter caught wind of what I was doing and called the house. Back then, the teenager whiz kid on the Internet was a hot angel. "Sorry, we're not interested," my Dad said before hanging up the phone.

Later that year, I was generously flown out to California and offered $2MM for my company. I rejected the offer and flew back home and went straight to the Chemistry class, telling the teacher I had a dentist appointment.

Soon we were hosting over 100 Million Polls and serving Billions of ads.

Over the years, I figured out how to hire people, build teams and run a business.

At some point, I read about something called an "Ad Server" and realized I needed one. An Ad Server is a piece of software that helps manage and schedule the billions of ads and track their performance. At the time the only player in town was a company called DoubleClick.

So I called them up. I think I was 15 years old at the time. "Excuse me, sir. I need an Ad Server". It was clear they did not want to talk to a 15-year-old, but eventually, I managed to pry a quote from someone. I don't remember what it was, but I remember thinking it was a ridiculous price.

I looked at the quote and thought, this isn't so hard, I'll just make my own ad server. So I built AdButler.com. I really just built it for our needs but figured, "Hey, this is a product, I'll release it to the world."

I was never really excited about advertising. AdButler was a tool that we used mostly for our own internal use.

In 2014 realized that Digital Advertising had solidified its role as the currency of creativity on the Internet. Almost everyone doing anything cool online was Ad-supported. From my favorite podcasters and bloggers to mega publishers like the Washington Post and The Guardian.

At conferences and events I attended, I would mention all the things I was doing, and as soon as I mentioned AdButler, their ears would perk up. "You know Advertising!? We need help with that." These were some of the people I most respected.

At the same time, Internet Advertising was evolving and began to present some of the most exciting and complicated engineering challenges in computer science. Advertising systems were handling insane volumes of traffic and data.

I had some ideas on how we could do better on both the business side and the engineering side, so I decided to jump into AdButler full time.

How are you doing today and what does the future look like?

AdButler has gone through the transition from a primarily SMB focused platform to a total market focus. In the past 12 months, our biggest deals grew from $50k/year deal sizes to $500k/year deals.

We are also in the process of transitioning from a single product company into a multiple product company. Something I think that every SaaS company should think about. Over the years, our platform evolved to be super horizontal, competing in 9 different verticals.

By splitting these 9 verticals into separate products, we will be able to communicate the value of each solution more clearly and realize increased pricing power allowing us to further grow the average deal size.

We also have some really fantastic technology in pipes. I believe that, in an increasingly privacy-focused future, contextual advertising is the future. Advertising that is focused on what you are doing or reading.

We are developing technology, using machine learning algorithms, that will read and categorize the content of a page the moment content is published. We will quickly understand that you are reading a sports article, specifically about golf and more specifically about the US Open.

Combined with our self-serve advertising portal, every website will be able to sell highly targeted advertising on its own platform in a privacy friendly way.

Through starting the business, have you learned anything particularly helpful or advantageous?

I have learned so much over the last 20 years, both what to do and what not to do. I'll share a few things in the order I think is most important.

1) Hire the right people. Your absolute number one job is, as Jim Collins puts, to "Get the right people on the bus." We have a top 10% hiring policy, and over the years, I have learned to be strict about it. This is especially true about tech people. Joel Spolsky says that the top 1 percent of the workforce can easily be 10 times as productive as the average developer. In my own experience, I have found this to be correct time and time again.

2) Work on your business, not in it. Think of your business as a machine. When something is broken or not working right, resist the urge to jump and play commando. Your job is to fix the machine, which might mean reassigning roles, hiring new people or implementing systems.

3) Beware of Savannah Syndrome. Imagine a giant herd of wildebeests migrating across the vast African Savanna. After trekking hundreds of miles, the herd encounters an alligator-infested river. The first few, understandably skittish, refuse to enter the water, holding themselves on the steep bank. As thousands more pile up behind them, they are pushed by sheer force of numbers into the river. The alligators have a feast killing many. After the slaughter, one lucky Wildebeest looks back at the blood-red river and thinks to himself, "I am so good at crossing that river! I am going to write a book about it."

You are far more likely to hear the stories of success and never hear from those that died in the river. Look for stories of failure and learn from other people's faults. There is often more to learn.

What have been the most influential books, podcasts, or other resources?

Good to Great by Jim Collins was the first business book I read and is still a great read. The Tao of Charlie Munger is a book I revisit often.

Predictably Irrational by Dan Ariely really helped me think about how people make choices and extend my model of pricing theory.

WaitButWhy is a great blog, specifically this post.

The Farnam Street Blog by Shane Parish.

Podcasts: Master's in Business and Invest like the Best.

Advice for other entrepreneurs who want to get started or are just starting out?

There is a tonne of advice out there and most of it is good advice so I will avoid repeating it but I will add one tidbit I don't hear a lot about.

Be wary of mediocre success. Mediocre success is way worse than failure. While failure will sting for a while mediocre success can torture you for years or decades.

I know a number of entrepreneurs who work like dogs, harder than they would if they just had a job, yet make less than they could work for someone else. They are working 80+ hours a week keeping their business on life support, taking on personal debt and making less than they would work 40 hours a week.

If this is you if you are just treading water, do something different. Do something risky, don't be afraid to push your business into failure. It's better than years of mediocre success.

I also know a number of entrepreneurs who failed hard and went on to create massively successful companies.

Where can we go to learn more?