When news broke in 2012 that David Cummings and Adam Blitzer sold Pardot to ExactTarget for a hefty sum, many in the Atlanta business community chalked it up to one of those nondescript, high-flying tech B2B startups that timed everything right. The dot-com bubble was still fresh off a generation of investors’ minds and Facebook still wasn’t making much money. The ambiguity of software and buzzy-tech words like “marketing automation” created a flash-in-the-pan, win-the-lottery type of perception. Of course, the storyline arc is no different from the gymnast who scores a perfect 10 in the Olympics or a golfer who wins The Masters and both become household names. Decades of work and refining a process culminate to one moment that can seem disconnected from the years of dedication and grind.
What few folks realize is Pardot was not the first business (or even 5th) started by Cummings. The first one was inspired by a prescient concept, mass concentration (an area of higher gravity on the moon making things more attracted to it), where he built shareware, arbitraged baseball cards, and built websites. Iterating through years of discovery, Hannon Hill started out of his Duke dorm room.
This is the story of everything learned from the ground floor and how one company or idea can lead to a better version in the future with hard work, vision, and a constantly iterated process.
ITERATION I: DEVELOPING THE BASICS
Jon Birdsong: When did you start your first business?
David Cummings: In 8th grade, I was on a travelling baseball team and one of my entrepreneur buddies gave me a book titled Teach Yourself C. This book taught me the basics of how to write code. I wrote a few programs and shareware apps. My first shareware app was called Statbook. This was for keeping track of your baseball and softball stats. Imagine a little app that replaced the physical book and did all your averages and record keeping and calculated it automatically.
Another one was called OnStartup.com.
JB: Do you still have that domain?
DC: Hah, no, but it would be a good one. OnStartups.com, with the “s” is the one Dharmesh Shah (of HubSpot) has—great site.
JB: That’s a real small world.
DC: Very true. OnStartup was software with programs for Spanish Word of the Day, Vocabulary word of the day, French word of the day, Famous quote of the day, Bible verse of the day, you know all those different categories that people would have on your desk that you’d tear off each day but it was an app version of that.
I sold about $1500 worth of shareware over two or so years.
JB: How did you sell it back then?
DC: They would download the free version from a shareware app store that included a two week free trial. If they wanted it for the year, they would send a check in the mail of $12.99 to my home address in Tallahassee. Then, immediately that customer would get an email with a code that unlocked a year’s worth of service.
JB: Were your parents wondering why the heck are these checks being sent to our house?
DC: Hah! Yes, I got about 2 a week, and for a 16 year old an extra $25 a week was a lot.
But the power of the internet didn’t hit me then.
I’m one of four brothers and my younger brother, pre-dot com boom, built a few websites — self-taught.
And at the time I was doing some sports card collectibles arbitrage. I would go on newsgroups and go on eBay. This was 1997. So in 1994, Major League Baseball went on strike. When the players went on strike, the value of all baseball cards just dropped off a cliff.
JB: What tough timing for the Braves because we were hot going into the ’94 season.
DC: Exactly. So by the time ’97 rolled around the baseball cards were trading about 25% – 50% of Beckett price — which was like the Kelly Blue Book for baseball cards. If someone in Seattle had a Braves card it was way less valuable to them than let’s say someone in Florida. The Marlins had just become an expansion team and because of TBS having national coverage, the Braves became the national team and America’s team.
I realized through the internet, I could buy baseball cards, mostly Atlanta Braves players and sell them to local sports cards dealers in North Florida for a good markup. Tallahassee, Pensacola, and Jacksonville were my normal go-to markets. So imagine a 17 year old kid coming in with a handful of popular Braves cards I bought for cheap and then selling them to local dealers. That was the aha moment of the reach and power of the internet.
I also felt like the math around buying $500 worth of cards and selling it for $1,000 was very tangible and therefore I could control my own destiny. I could do more work and make $500 instead of waiting for a check or two every week.
It was around this time the dot-com boom was in full effect. Because I built shareware software, and I built a little website for the sports card business, then it was just boomtime for small businesses that wanted websites built. By senior year, I was charging $50 / hour building websites for people. Customers included a law firm in Tallahassee, an auto parts accessory store, and a few other small businesses in Florida.
JB: All HTML?
DC: Yeah, HTML, a little bit of Dreamweaver, and a little bit of Frontpage.
JB: Are you making good grades at the time?
DC: Yeah, all A’s.
JB: At the time you’re starting to look at Universities, why Duke?
DC: On the school front, my dad and older brother went to Carolina, so I always heard good things about that. What drove my decision was the US News and World Report at the time had Duke ranked number 1 in the South, which it still is 20 years later. At the time it was 3rd in the country. It goes back and forth from 3rd to 9th, for undergrad. So I read about it and figured it was good plus my brother was 8 miles down the road, so it was a good option.
JB: When you go to Duke, I assume you’re still able to build websites for folks at $50 an hour?
DC: Yes. Building websites for $50 an hour in the late nineties as a teenager was crazy! $3.15 an hour was the minimum wage. Publix was the go-to in Tallahassee for your high school job.
JB: What did you spend the money on?
DC: Reinvested it in the business. Gotta have the latest laptop, flat screen, etc.
We both start laughing.
All the money went into the latest and greatest hardware updates and improvements.
ITERATION II: ENTREPRENEURSHIP AT DUKE UNIVERSITY
DC: In Fall of 1998, I head to Duke and start building websites there. Freshman and sophomore year, I continue to do the web design business, which included the Duke Computer Science Department, the Psychology department, different sororities around campus, and a non-profit on campus. There were many!
JB: How many pages are we talking about?
DC: Less about the number of pages and more on design. I would do full-on Photoshop design and the whole nine yards. I remember, my freshman year roommate was this crazy smart guy who went to the North Carolina School of Science and Math, which is where Hubert Liu went, which is an extremely good school.
So a few years prior, I had been building dynamic apps but the websites had all been static with HTML and Photoshop. My freshman year roommate shows up to campus running a Linux box in our dorm room. The Linux server which is widely popular today, but back in 1998, it was unheard of to be running a Linux box.
And he’s running this Linux box and writing these Perl scripts which, at the time was the best way to build dynamic websites, and here I am coming in with more of the front-end developer skill set and my college roommate has all the back and devops system admin skills. It was a perfect world.
With our match of complementary skills, opportunities started appearing. One of them being twenty years ago, you’d buy a textbook from the campus bookstore for $200 and end of the semester you’d sell it back for $20.
JB: This sham is still going on today.
DC: Yes, still going on today!
I got through the end of the first semester of freshman year and saw that experience, so immediately at the start of next semester, my roommate and I start hacking away at a webapp for students on campus to sell directly to each other.
We launched that second semester before the end of the school year so that students could buy and sell from each other before going off to summer. That was my first serious go at a dynamic web application.
It was really a marketplace.
JB: And it worked?
DC: We made little flyers and focused just on freshman because we were on freshman campus – which was around 1,600 students in the freshman class. We went and put a flyer on every door on the freshman campus, which was 800 or so dorm rooms.
JB: Who was more the marketer of the two?
DC: He was all back-end developer and I was all front-end, which included everything from design to marketing, including putting flyers on every door.
So we blanket the freshman campus with this, Textbooks.ml.org, and I immediately get contacted by the Duke bookstore that wants to talk about it.
So I get called into the bookstore and talk to the management there. In our conversation, he’s going through why they have to charge what they charge and why they have to buy back books at the price point they buy them back at and I’m like, “Great, you keep doing what you’re doing and I’ll keep doing what I’m doing.”
I thought it was fun to be asked to come to learn about the operation.
It was clear, just like with any local monopoly, they were just trying to defend their turf.
Lots of lessons learned there and that was freshman year.
JB: Did you keep it going throughout school?
DC: There was no way to monetize it so we left it up there for people to use but without putting the flyers under the doors near the end of every semester, it wasn’t going to work because students didn’t have a reason to come back, there was no daily active user component of it.
We abandoned it because we couldn’t figure out how to make money. We could throw ads on it, but there was such little traffic. We could have taken it to other schools, but we hadn’t made any money at one school so why go to others?
That was my first serious, dynamic webapp and micro-marketing campaign.
JB: How does that experience lead to starting Hannon Hill?
DC: All these steps, from building installed Windows shareware, to static websites, and now dynamic webapps led me to another idea which started sophomore year. The idea, and remember this was 20 years ago, was an online way to rate your professors. At the time, it didn’t exist.
Ratemyprofessor.com didn’t exist yet.
I had all these ideas from the pains of students, for example, getting the list of courses and professors and not knowing what they were like was a very big one. Another one, and I didn’t start this one, was basically a message board for students to talk about whatever they wanted. There was no Facebook at the time, there was no Twitter, no Myspace, it was before all of those.
Someone had put a Duke specific message board online for people to talk about stuff and there was a section that was anonymous which was a precursor to YikYak and that was obviously the spiciest and most controversial section.
Then there was the headache of laundry. Just like 2ULaundry today. So I put together a site called DevilLaundry.com and I partnered with a local laundromat, the closest one to campus, which was called White Star Laundry.
I Iterated on what I’d learned before—built the website, partnered with White Star, put the flyers under the dorm room doors again, and got the students signed up for a weekly service. So they’d go on the web app and note: I want my laundry to be picked up Saturday morning at 10:00 a.m. and be ready for pickup Sunday afternoon, and they could pay with credit card.
I did a similar thing for the coffee shop on campus called Alpine Bagels. I built them a website so they could take orders online and offer delivery of coffee, bagels, and fresh squeezed orange juice (which was their best seller). This is probably the year 2000, so way before delivery was popular.
All of these were decent ideas, but back then scaling something like that took a lot of money and resources and the monetization strategy was still unproven for many of them.
Two major takeaways from all this work was honing the ideas that stemmed from problems that I or my friends kept coming across, and also crafting my skills at building web apps, websites, marketing, and going to market with partnerships and selling them on what ‘could be’.
JB: Did you make any money off them?
DC: In hindsight, I made a tiny amount of money on them but none of them took off for a variety of reasons. But with every one of them, I learned something and I got a little bit better at the technology and functions to create.
That lead to junior year…
JB: Before we get to junior year, what are you doing in between summers?
DC: After my freshman year, I was the only freshman to get a full internship at IBM, writing code.
JB: Wow.
DC: I got paid $20 an hour—full-time—to write code for their Tivoli product, which is basically a single sign on type system for big enterprises. I worked in Research Triangle Park and lived at my brother’s apartment who was in Carolina at the time. After the summer, it was abundantly clear I was never going to work for a big corporation. Never.
We both start laughing
That was freshman year, and the summer of sophomore year, I went overseas and studied at the London School of Economics and then back packed around Europe with my best friend. All in all, I spent 10 weeks in Europe that summer.
ITERATION III: STARTING HANNON HILL — RIGHT PRODUCT, WRONG TIMING
JB: You return back from your trip and immediately start Hannon Hill?
DC: The first line of code for Hannon Hill was in my last week of my first semester junior year so call it December of 2000.
The idea with Hannon Hill was, at the time, I had built dozens of websites which was the main money maker. When I would hand over the website to the customer and it was shiny and new, the customer thought it was great, but when they wanted to change it or update it, back then there was no WordPress, no Squarespace, no Wix, no Joomla, no Acquia. Nothing existed to make it really easy to update a website.
This was December of 2000. So I started writing some code and then January 2001, I pitched a professor that I developed a good relationship with on this idea for Hannon Hill making software to make it easy to update and maintain a website for non-technical people.
JB: How did the professor respond to the pitch?
He was the prototypical eccentric, brilliant professor. He wore all black everyday and told himself when he became a full professor, he would never cut his hair again. So he had a massive beard, a long ponytail, probably weighed 300 lbs., but the most crazy smart and crazy personable professor you could think of. He became famous in the academic world for using technology for foreign language learning.
He believed in me personally.
When I pitched him, I asked him for $20K so I could build software to update websites.
The name of the company was obviously called Hannon Hill and the name of the CMS (content management system) was called Super Update.
He said, “Okay, I’m in. I’ll write you a check for 20 grand.”
JB: Did you send him a pitch deck?
DC: Nope, it was more simple than that. I told him what I wanted to build and how I was going to build it. I was going to hire a few classmates to help me write the code for software to make it easy to update websites.
When he said yes, I started the process of finding a one room office that was close to campus which ended up being on 9th Street in the Couch Building. Rent was $300 a month for that one room.
JB: Did you know who the customers would be?
DC: They would be the ones who I already built websites for and were constantly asking me to update their website.
JB: Why didn’t you just keep building websites?
DC: I realized a while ago there wasn’t scalability in just being a consulting agency. It also didn’t scratch the itch of building a large, meaningful business.
JB: How many customers does Hannon Hill have today?
DC: Over 300 colleges and universities across the country and even outside the country.
JB: How did it morph from CMS for anybody to universities?
DC: It started out as a CMS for anybody because I didn’t know any better. It morphed to CMS for small businesses and iterated to CMS for schools and universities.
Long story short, I started to sell Hannon Hill into agencies but I realized the agencies didn’t mind charging their clients $125/hour to update their website. It took me a while to realize that, but from there I went direct to the customers.
I took a leave of absence from Duke first semester senior year to sell our product that we were building. It was at that time 9/11 happened and the biggest conference of the year called Internet World, which was held in New York, got pushed back from September to December.
The same friends who helped me build the software finished up exams and flew up to meet me for the conference. I remember I borrowed my brother’s green, four-door Buick Skylark, filled it with the 21-inch CRT monitors with full desktop computers, and by myself ,drove it New York City. My buddies met me after they flew up…
JB: How many paying customers did you have the this time?
DC: Three or four.
JB: Amazing. So big faith in the product and market?
It wasn’t even faith, it was just being blissfully ignorant. The strategy was “just keep doing stuff and keep moving.”
When I got to New York and started unloading all the equipment, I remember trying to start the car again and it didn’t start! Looking back, I remember that was a massive pain. Anyways, we’re in the Javits Center, and the booth cost $5000. Fortune would have it, two rows over, there was a company based out of Texas called Globalscape which at the time had the world’s most widely used FTP product and had 1.2 million paying customers, small cap, publicly traded, and they were looking for a content management system to upsell and cross-sell their existing FTP customers.
JB: No way.
DC: We were in the right place at the right time. They had a biz dev team member looking for partnerships and the like. We left keeping the conversations going.
But at the time, since the business wasn’t taking off, I went back to school part-time which was my second semester senior year.
JB: Your year is graduating at this point?
DC: Correct and I had applied to graduate early before I took time off and since I was going back part time, I could now apply my AP credits from high school and finish school on time while effectively taking only 3 and 1 quarter year of classes.
JB: Was your degree important to you?
DC: It wasn’t important, but I just figured I’d check that box since I was so close.
My senior year spring break, a bunch of my buddies were going down to Miami. I ended up flying to San Antonio which is where Globalscape is headquartered. I spent the whole spring break in Texas, training everyone on the Content Management team at Globalscape how to use Hannon Hill’s Super Update that was reskinned under their own brand and they were calling it PureCMS.
We ended up closing the deal with Globalscape which included a prepay of $200,000 in royalties. The final deal was they get 80% and we 20% of everything they sold, but they would sell it, brand it, support it and more. No one would know Hannon Hill was behind the software.
This $200K was non-dilutive cash in the bank regardless of whether they sold 10 or 10 million.
ITERATION IV: SAME NAME, NEW PRODUCT ITERATION
At that point, outside of the Globalscape deal, I was discouraged enough trying to sell a $30/month SaaS product to small businesses through their web design agencies.
JB: What did you do with the $200K?
DC: I re-hired the classmates from the previous summer to build something more substantial. We didn’t want to do the small business product any more…
JB: Self sign on hadn’t caught on yet?
DC: SaaS hadn’t caught on yet.
We both start laughing
SaaS really came into its own in ’07 or ’08. The great recession really made it so big IT spend went out the window. The great recession really helped the adoption of SaaS because you don’t want to pay $100K upfront, you want to pay $2,000/month forever.
This was way before primetime SaaS.
So I take the $200K with my classmates, and start from scratch a new product from the ground up. That product is called Cascade Server which is the product Hannon Hill still sells today and it’s been wildly successful.
KB: What type of time commitment are you putting into this?
DC: 16 hour days, all day, 7 days a week.
Cascade Server was a new content management system built from the ground up, but this was built in Java where the previous one was built in PHP, and Cascade Server was an installed enterprise app, it was not SaaS. So this is heavy duty, powerful software you install on your own server.
JB: Why did you decide to move the company to Atlanta?
DC: At the time, my biggest web design customer was also my biggest Super Update customer. They were a private equity hotel group in Atlanta. I reached out to my contact who was our customer and said, “We’re looking to make a move to Atlanta.” I had always liked Atlanta because it was the city I was used to going to growing up in North Florida. The Braves, airport, Six Flags, shopping, etc. All of these are reasons I visited Atlanta a bunch growing up.
When I called up our customer, he said they had two empty rooms in the back of their office that we could use. At the time, he was on the 11th floor of the Monarch Tower, which is attached to the Ritz-Carlton and looks out over Lenox and Phipps, class A building. That building was built in ’98 so it was only four years old and there was an apartment complex right across the street. Once I saw all of that, I was sold.
JB: When did you get your first paying customer for the Cascade Server product?
DC: We started writing the new product in Java from scratch in May of 2002 and it was really early 2003. At this time, I still didn’t have a clue about customer acquisition and go-to-market.
JB: When did it finally hit you that picking up the phone was important?
DC: I did some cold calling for Super Update in Durham to design agencies and that failed for a lot of different reasons.
But in 2003, we had a little bit of money in the bank, I had hired an amazing senior software engineer I found through Craigslist and then Summer of 2003, I got a really smart Emory intern—a CS major—and then another other buddy from Emory who was on the product management side. It was really us three and the intern for most of 2003.
JB: Somewhere in 2003 is when you got your first customer?
DC: Correct, we got a customer in 2003, which wasn’t that big at the time. A few customers started using us for their intranet. I started doing some Google Adwords ads for content management and I got Eckerd College down in Florida to sign up.
JB: How much were those deals?
DC: They were around $20K, paid upfront.
JB: Did you have to go down to Florida to implement?
DC: I did not. Sold it over the phone.
Over the next few months, we landed a few more customers through calling and Google Adwords.
JB: At this point, did you feel like you were on to something?
DC: Once we got to $50K of product revenue for a customer, I felt like we were on to something. Raising prices was a big breakthrough.
Then we had a little bit of cash and hired two Emory interns to be cold callers and got a referral from a family friend to hire Blaine Herman, who now has gone off to start his own company which is doing very well.
Since two of the five first paying customers were colleges, we decided to double down on colleges and universities. We went to Barnes and Nobles in Buckhead and bought a massive book with all of them listed out: two year, four year, public and private, all listed out. There is 4,160 or there was at the time and we all started dialing for dollars and it worked amazingly well.
I did about 1,000 demos. 50% of my time was doing demos set up by the team.
It was good fortune the ratios worked around ACV, CAC and other SaaS metrics we all know well today. But at the time, we weren’t paying attention to that, we were just working as hard as possible and as smart or semi-smart as possible.
We were blanketing colleges all over the United States.
JB: What did your annual revenue numbers look like?
DC: 2003 was probably $20K-$30K of revenue. In 2004, call it $200K in revenue maybe $300K. In 2005 we went to $1.2 Million in revenue. In 2006, we did $2.4 Million.
The beautiful thing about this revenue was it was all prepaid and upfront which helped us reinvest into the business quicker.
JB: Did you ever think about raising money?
DC: I talked to ATA (Atlanta Technology Angels) and multiple angels in the area and they had no interest. This was ’05 and ’06, the real estate market was booming and at the time, I was 25. So you have this 25 year old with a million dollar revenue, enterprise software installed content management business—it just wasn’t interesting to people.
Today, Hannon Hill continues to thrive out of the Atlanta Tech Village with a team of 30 people. Over the last 18 years, Hannon Hill has sold tens of millions of dollars of software and continues to grow every year.
Hannon Hill is the story that rarely gets told. People hear the Pardot story, the Atlanta Tech Village story, and more, but Hannon Hill was the spark that set it off.
After Cummings exited Pardot, $500,000 was donated to Duke in honor of Professor Frank Borchardt, who passed away in 2007, which turns into $20K a year forever to the top undergraduate entrepreneur as a gift in his name. Duke entrepreneurship leaders select the top startup.
Every year, a Duke undergrad gets the equivalent of what he received from Frank and non-dilutive.
ITERATION V: PARDOT AND BEYOND
JB: You’re growing a great business that’s profitable and doing well. Why did you start Pardot?
DC: At the time I started thinking about Pardot, which was around 2006, I had a much bigger team on the Hannon Hill side and I was able to spend a lot more time on sales and marketing. As I spent more time on that side of the business, it soon became clear that we could do more with the help of more technology on the sales and marketing front. I felt there was something there in that market—using technology for a B2B marketer—because I was a B2B marketer looking to be more efficient.
A few other reasons included, I felt like a pure recurring business was going to be the future. It was more predictable revenue wise vs. the super lumpy revenue that comes in with on-premise, installed software which created a feast or famine revenue model. I wanted to find a recurring model that was most predictable and compounding.
Building and exiting Pardot with Adam Blitzer provided the resources to support so many entrepreneurs that I respect and admire today.
JB: Alright, before we finish this interview, I want to ask some random, off the cuff questions. I’ve realized working with you over the past year, you’re constantly thinking of new ideas. Can you describe to us, how can one can constantly think about new ideas?
Every conversation I have, every article or book I read, every talk or podcast I listen to, I think about businesses to start. The short answer is reading, learning, and talking with as many entrepreneurs as possible. Always being in motion. It’s not intentional — it just happens naturally for me.
JB: After you sold Pardot, and that story has been told by many, you bought an old, stodgy building in Buckhead immediately afterwards and almost everyone in the real estate world thought it was crazy. The idea of kicking out current tenants to bring in a bunch of high risk startups with no operating history, no balance sheet, and no leases, into the building. Fast forward six years later, and there are over 1,000 members, 300 startups, and the economics of the building is working well.
What did you see that few other did not?
DC: The traditional real estate investor did not see the headaches of an entrepreneur trying to get decent office space. They couldn’t feel that pain from the entrepreneur’s perspective. Their whole world was the bank that’s going to finance this needed creditworthy tenants with long term leases and more. They couldn’t see there was a whole segment of the market, a small but meaningful market, that was two entrepreneurs and an idea needed space. And those two entrepreneurs with an idea can grow to 10 or 20 or 400 employees with time.
JB: This was three years before WeWork came to Atlanta.
DC: Exactly. But this wasn’t on their radar because it wasn’t worth their time.
JB: Give us your thoughts on SaaS 10 years ago, SaaS today, and SaaS 10 year from now.
DC: I tried to sell SaaS in 2001 and it was too early and it did not work. The market wasn’t ready. Customers wanted to install software and own it. From 2007 to 2012, through the worst recession in many decades, there was still friction to implementation because many customers still had their legacy product, but it was booming compared to the years before, but it was still a really small market.
Today, it’s still booming but now the market has compounded for 10 years and so now it’s a medium size market. So it’s gone from tiny, to small, to medium, and the next 10 years is going to be taking all the ones that are winnering today and then thinly slicing those segments that can be better served. For example, if you have 4 billion dollars of revenue from Salesforce.com software just in CRM, you can imagine there is going to be 20 other CRM’s that solve specific CRM problem for different segments of that market, but because it’s focused on that segment, it’s going to be better faster, cheaper. Even now the incumbents are so big, they can iterate fast enough and a small percentage of their business is a meaningful company.
Collectively, those companies will be tens of billions of dollars in revenue and the Southeast a great place to build those types of businesses.
JB: What gets you excited over the next 10 years?
DC: Any way to build businesses that fulfill the entrepreneur’s dreams and ambitions. Building companies that solve a problem and serve a real need is so exciting to be part of.
ITERATION VI: THE ATLANTA VENTURES STUDIO
Ready to be written…
Story by Jon Birdsong