7. Kamakshi Sivaramakrishnan – raising the Drawbridge

Kamakshi Sivaramakrishnan is on the product team at Microsoft’s LinkedIn, which acquired her company Drawbridge in 2019. She founded Drawbridge in 2011 and was its well-known CEO for eight years, making the case — largely successfully — for the validity of probabilistic methods of determining identity in marketing and advertising, where 100% accuracy is not only not requisite but almost never possible. Drawbridge was for some years the ad industry’s leading solution for determining and managing cross-device identity — i.e., determining that different devices belong the same person, household of account using some complicated math.

Before founding Drawbridge, Siviramakrishnan was Lead Scientist at AdMob, a mobile ad platform that served mobile app developers (starting in the WAP era of m.dot apps); it was acquired by Google in 2009. She has a Ph.D. from Stanford in Information Theory and Algorithms, making her the first “doctor” we’ve had on the show. She was born and raised in India in the former Bombay (now Mumbai) and is the first female Ph.D. in her family.

In this rapid-fire chat, Kamakshi shares about her childhood and her ambition to be a mathematician, her academic dreams and how they were sidelined by a successful courtship by AdMob, which started her journey in ad tech. Along the way, she shares what it’s like to shop for real estate in S.F. and how she managed to sell her company to Microsoft and give birth to her child at the same time.

6. Cory Treffiletti – agency guy to BlueKai, Oracle and beyond

Cory Treffiletti was an ad major at Syracuse and one of the “lucky ones” — as he says — who know what they want to be early on. Moving to NYC, he joined an agency and was given digital planning responsibilities for major brands such as BMW and CVS just because he’d “been on the Internet.” It was 1995. He later pioneered media planning and testing techniques at agencies such as i-Traffic, Freestyle Interactive and Carat Fusion, in the days when home-page takeovers and roadblocks on Yahoo were the norm for big campaigns. Moving to the West Coast, he lived through the dot-com boom and bust — “It was really fun!” — and eventually founded a startup advisory/capital firm called Catalyst S+F, where he played short-term CMO roles for startups such as Kenshoo, Turn, MediaMath and RedAril. He joined BlueKai as CMO in 2012 and helped define the DMP space, later becoming CMO of Oracle Data Cloud after Oracle acquired BlueKai in 2014. Cory is currently SVP of Marketing at fintech platform FIS. A longtime columnist for MediaPost, he’s also author of the 2012 book Internet Ad Pioneers, a collection of in-depth interview with notable women and men from the first 20 years of digital media.

In this wide-ranging interview, Cory takes Jill and Marty through his journey starting in 1995 as a tyro at the dawn of the Internet, through the dot-com boom and bust, his time on-stage with Primus … to CMO-dom and Oracle Data Cloud … meanwhile providing a lucid explanation of exactly what a Data Management Platform (DMP) is, anyway, and a pointer to one of the all-time great pieces of content marketing.

5. Paul Bannister – gamer turned prophet for the publisher

Paul Bannister started life as an intrepid video gamer who turned his passion for Nintendo into an online publication that — in the early pre-history of 1995 — was one of the first to sell an ad on the internet. It cost $400 for an arbitrary-sized (but small) highly pixelated ad unit with a lot of text. His career as a publisher was launched. Years later, he joined Andrew Shue (Melrose Place, soccer star) and others in co-founding CafeMom, an early and successful example of a social network-content site. CafeMom was ad-supported from the beginning, and from his perch at the ever-growing network Bannister saw the development of ad servers, ad networks, SSPs, DSPs and header bidding. Today he is Chief Strategy Officer for CafeMedia, a company based largely on the acquisition of AdThrive in 2018. CafeMedia basically outsources publishers’ back offices and revenue ops so they can focus on creating content and is particularly strong in travel, hospitality and leisure. In this wide-ranging episode, Bannister trains his widely-admired prophetic abilities on the history of publishing tech, the future of the cookie, and why there may be some reasons for optimism looking toward 2022.

4. Joanna O’Connell – evangelist of the trading desk

Joanna O’Connell is a well-known industry pundit and V.P. Principal Analyst at Forrester, where she covers advertising technology, a field she has graced for more than twenty years. Starting as a celebrity-loving production assistant at CNN, she joined Avenue A as a media planner in the late 1990’s in New York City, left to join a fraught Miramax for three years after a twenty-something crisis of meaning, and returned to Avenue A as it transformed into Razorfish and took the mathematical road to media optimization. A chance meeting in an East Village nightclub around 2007 started her on a continuing role as programmatic evangelist and led her to co-found ATOM Systems, the first big agency trading desk (build on MediaMath, then Invite Media). Pro-programmatic evangelism defined her work at Forrester and later AdExchanger, where she was chief of research until joining MediaMath itself as CMO for two years (2015-17). In this interview, Joanna talks to Jill and Marty about her career path, her love of “nerdy” media buying, and her mad-scientist father who turned a horse barn in Upstate New York into an inventor’s lair.

3. Joe Zawadzki – the original (Media)math man

Joe Zawadzki is CEO of MediaMath, ad tech pioneer and noted investor, mentor and visionary. A floppy-haired, Tesla-driving, intensely affable Harvard grad, he came to NYC in the late ’90s and joined a real estate firm as an analyst. Rapidly, his interest turned to e-commerce and personalized shopping apps using ASP. (In this episode, he reminds us that in the pre-SaaS stone age, start-ups had to buy their own servers and security systems.) Through the time of the dot-com boom and bust, he bravely founded Poindexter, a Silicon Alley firm that provided technology to agencies. Poindexter was originally focused on website personalization and optimization, and customers such as AOL and American Express asked them to deliver better prospects in the first place (aka, ad tech). Zawadzki adapted optimization to ad targeting and serving, was involved in the ideation of Right Media — his own ‘origin story’ and cautionary board-handling tale is dramatically told here — and saw Poindexter rebranded to [x+1]. In 2007, he founded MediaMath, a leading DSP, which continues to rock and roll today. He’s also a co-founder of MathCapital, an investment fund.

2. Zach Rodgers – the voice of AdExchanger talks

Zach Rodgers is the editorial director of the mighty AdExchanger, founded by John Ebbert 12 years ago and still representing the pinnacle of ad tech journalism. In fact, we’re lucky as an industry to have such a voice. He joined the editorial team more than nine years ago and has been a key influence on its integrity, quality and growth – as a writer, recognizer of talent (often quite young talent), event shepherd, manager, editor, source wrangler, MC and raconteur. In the latter role, he’s hosted the 226 episode-strong and counting AdExchanger Talks podcast, which has seen just about every ad tech influencer shuffle behind its mike. A part-time vegan and all-around mensch, Zach started at ClickZ and other trade pubs before entering ad tech history & legend. In this episode, he talks about his childhood dreams (many), his initial opinion of advertising (not so hot), how he met John Ebbert, AdExchanger’s legendary first event at the New York Historical Society, his pursuit of Brian O’Kelley, gender bias at ad tech events, and the future of his storied pub. (Also check out Zach’s side hustle, a podcast called Beaconites about his beloved hometown of Beacon, NY aka ‘the New Brooklyn‘.)

1. Eric Franchi – the many overtones of Undertone

Eric Franchi is tri-state native and son of public school teachers. He made his way to About.com as a hard-charging sales rep during the late-90’s dot com boom, selling digital ads to brands who were about to feel the pressure of a reeling post-2001 economy. With a buddy from About.com, he went on to found a company that was eventually called Undertone. He and David Cassidy started as a two-man shop with a fax machine in a Silicon Alley warehouse, buying ads for Orbitz and Netflix in the days when they were soliciting subscribers for their mail-order DVD business. And then he discovered pop-unders — a short-lived but high-impact ad unit that was richly interactive and (in time) richly overused. Then came browser pop-blockers and the refashioning of Undertone into a premium ad network with about 300 employees when it was sold to Perion in 2015. Franchi is currently co-founder and principal at MathCapital, an early-stage venture fund that focuses on ad tech visionaries.

“It Wasn’t a Job, It Was a Cult!”

The Paleo Ad Tech podcast was inspired in part by an ‘oral history’ project one of us (Marty) worked on for a couple of years while he was a Gartner analyst, covering ad tech. (They are archived here.) He interviewed a number of the key players in the industry about their experiences and committed write-ups to the Gartner blog. One of them (an interview with the mysterious ‘Dr. Boris,’ behind-the-scenes builder of much of the early ad tech infrastructure, starting with Right Media) was Gartner’s most popular blog of that year. The series ended in 2018 when Marty left Gartner — but not before the following ‘definitive’ oral history of DoubleClick ran in AdExchanger, on October 29, 2018. (Marty had to publish it anonymously due to his new employers’ PR code.)

‘It Wasn’t A Job, It Was A Cult!’: The Definitive Oral History Of DoubleClick

Suddenly, last summer, Google announced it was erasing the DoubleClick name from its product portfolio. So ends 23 years of ad tech myth and legend that coincides with the ad-supported internet itself.

DoubleClick rode the dot-com boom to heights of overcapitalization and came crashing down with everyone else in 2000. It created New York’s Silicon Alley, had outrageous parties and heavenly bills and saw just about every future ad tech luminary pass through its cult.

The iconic sign that loomed over Broadway and 22nd Street during the days of the boom defined an era:

“DOUBLECLICK WELCOMES YOU TO SILICON ALLEY.”

It was an ad network and ad server before anyone knew what those were. It provided a sales team and infrastructure for advertisers to buy ads from multiple publishers in a single location, target them using criteria such as location or time of day, insert them into web pages automatically and get reports. It has quietly dominated ad serving since it started and thrived as a brand since being acquired by Google in 2007.

Success is obvious in reverse. But in the summer of 1995, two engineers named Kevin “KO” O’Connor and Dwight Merriman were sitting in O’Connor’s 2,500-square-foot basement in Alpharetta, Ga., looking for something to do. They had cashed out a networking startup in Ohio and moved south to raise families.

Kevin O’Connor, co-founder: Dwight and I came up with hundreds of ideas and narrowed them down. The original concept was a network of publications, like a cable subscription model. Then Dwight looked at how media monetizes and said advertising could be bigger. We flipped the idea into a network of advertisers.

David Gwozdz, sales director: Kevin called me and said, “I understand you’re a good ad sales guy.” I went over to his house to tell him no thanks – and ended up taking the job. Kevin met my wife, and the first thing he said was, “I’m going to make your life hell.” We worked seven days a week, morning to night.

O’Connor: Dwight was coding, and I was learning about advertising and direct marketing from textbooks. I was going to the library reading Ad Age and Adweek to see if we had any competitors.

Gwozdz: At that time, the ad server was an ISDN line and a 486 PC on Dwight’s desk with the cover off, because it kept overheating.

O’Connor: I read in an article that an agency called Poppe Tyson was doing the same thing. I thought, “Holy crap!” and called a guy there named Dave Carlick.

David Carlick, EVP, Poppe Tyson: Our interactive group was called DoubleClick. The name was invented by [Poppe Tyson CEO] Fergus O’Daly. It wasn’t a good name. It’s an old Apple term. A click on a URL is a single click, not a double click.

O’Connor: It turned out they didn’t have any tech yet. It was the classic, “Let’s launch an idea and see how the market reacts.” We were just completing the product, and they had a team of four selling ads. We decided we should work together.

Carlick: There was a long discussion about whether it should be based in New York or Silicon Valley. We decided on New York and started the whole Silicon Alley thing. The story is that when Kevin and his wife saw the tiny apartment they would have to live in, they cried.

After a tentative start, the company started to, um, click but banged into sales teams from websites like Netscape and Excite, which were Poppe Tyson’s major clients. So Fergus O’Daly decided to spin off DoubleClick into a separate subsidiary with Kevin O’Connor as CEO.

O’Connor: We were 50% owned by an ad agency. No one liked that. We tried to sell the company to someone else. There was a deal with Yahoo back in 1996 – we wanted $100 million and they offered $95 million. They wouldn’t budge.

Kevin Ryan, CFO (later CEO): I was thinking about starting an ad network, so I met with Kevin and Dwight. I thought they were very good – they had a six-month head start. So I became employee No. 12. We grew very quickly. Four years later, we had 2,000 employees in 25 countries.

Gwozdz: I was still in Atlanta, in a little office usually by myself. I remember I went to a sales meeting in Colorado Springs, and there were 200 people in the room. I didn’t understand how big this was getting.

O’Connor: Everything started to go through the roof – a lot of revenue, with hefty margins. We needed capital. The mantra was “get big fast.”

Gwozdz: We started picking up pub after pub. We were an upstart with really good tech no one else had. We could target by location, time of day, exclusion. People would light up. They loved the reporting.

Bill Wise, director, financial planning (later VP/GM): KO would get up in these meetings and he was so aggressive – we loved it. He said, “We are the company that is changing advertising and media. We won’t stop until we have world domination.” It wasn’t a job, it was a cult – a feel-good cult.

DoubleClick went public in 1998 on the Nasdaq exchange under the ticker symbol DCLK. It was the beginning of the bubble’s last bleat. The Nasdaq rose 86% in 1999. It soared 5x from 1,000 to 5,000 in the five years before 2000. There were 15-20 dot-com parties every week in San Francisco. About 70% of digital ad dollars came from VC-funded dot-coms. Of course, few people suspected what was coming. They were just trying to find a place to sit.

Wise: There was a temp who was working under [VP of sales] Wenda Millard – a great kid – he was sitting in the hallway. One day somebody put a “Kids in the Hall” poster up behind his desk and crossed out the “s.” He was known as the “Kid in the Hall.”

Brad Bender, director of product management (later VP): They put me where they’d moved a copier out. It was the world’s smallest desk. I was the “Kid in the Hall.” The executive team was right around the corner from me. I remember Kevin O’Connor came by with clients and said, “Look, we’re growing so fast, we even have a ‘Kid in the Hall.’”

Wise: Eventually we maxed out the space and threatened Manhattan that we were going to move to Jersey City. They ended up giving us tons of tax credits to keep us, since we were the founders of Silicon Alley. We built out the space on 33rd Street and 10th Avenue.

Scott Knoll, director of business development (later VP/GM): 33rd Street was a brand-new building. It had a climbing wall, a multimillion-dollar basketball court. We had to build a special platform to muffle the noise for the tenant under us. It was spectacular. There was a view of the Hudson River. Bleachers. Fiberglass backboards, a scoreboard and lighting.

Wise: DoubleClick had the best parties. KO was at sales conferences playing guts and taking stock options as currency.

Ryan: We went all-out at Halloween. Everyone came to the office in costume. One year I showed up as a mummy. I had trouble getting a taxi and I couldn’t get into the building. I had to remind the manager that he knew me. Then 200 of us marched up Fifth Avenue at 10 a.m. in our costumes.

Then the company got a real scare. It decided to merge with Abacus in 1999 for $1.7 billion. Abacus was a consumer data co-op that tracked the catalog buying habits of most US households. Trouble was, it used personally identifiable information, which DoubleClick proposed to map to its (anonymous) cookies.

O’Connor: Abacus was a great company that had a tremendous amount of offline purchase data. That acquisition was about ways to combine offline and online consumer purchase data. We weren’t wrong about it. But we came up against privacy concerns.

Wise: USA Today published a story that DoubleClick was going to try to marry online and offline data. It got picked up by other publications and got a lot of attention.

Knoll: Abacus was frustrating. It was a situation where the influence of media was major but they didn’t have the story right. Nothing had actually been done yet.

Wise: KO took the fall, as a great leader does. He cared more about the company than his own legacy. It was sad how it went down.

O’Connor: I can’t say Abacus was overblown. It is what it is. People were afraid. I left [in July 2000] because I couldn’t see a path to where I could enjoy my job. I’m a product guy and an engineer. I like building things. I was a reluctant CEO. In the end my job was just dealing with politicians, the media and lawyers. I wasn’t good at it. I was burned out. Fortunately, Kevin Ryan loved dealing with those folks, so he stepped up from president to CEO.

What went up came down. In the 30 months after the market peaked on March 10, 2000, the Nasdaq fell 78%. Between 2000 and 2001, the Super Bowl went from airing 16 dot-com spots to three, two of which were job-hunting sites. By one account, CPMs for banner ads fell from $50 to $5. More people were leaving the Bay Area in panic than arriving with hope in their hearts.

Ryan: Most of our publisher clients went bankrupt. It was Pets.com times 100. Those were our clients. We did seven rounds of layoffs and went from 2,000 to 1,000 people. We turned over the entire management team. A lot of them couldn’t do the sixth round and still feel good about it.

Bender: It was a challenging time. It wasn’t clear when we were going to be done with the cuts. A lot of people I valued as colleagues and friends weren’t around after that period. But it was a healthy retrenchment.

Ryan: We lost 30% of revenue but eliminated 50% of cost. So by 2004-05, we were actually profitable.

Ari Paparo, VP, rich media: When I joined [in 2004], there was no structured onboarding. My boss quit on my third day. The layoffs were done, but it was just a crappy company. It was dysfunctional, engineering products that were incredibly bad.

Ryan: In 2005, our valuation was very low, just like everyone. The board felt we were undervalued, so they hired Lazard and interviewed 50 buyers.

Paparo: The company announced publicly it was going to look at strategic alternatives nine or 10 months before [SF-based private equity firm] Hellman & Friedman closed. Even Google bid a fraction of what they later paid.

O’Connor: Google’s perception was that we were a sales company. They thought tech could only be built in Silicon Valley. They were wrong – as they realized later.

Ryan: DoubleClick sold at a 50% premium to the stock price. Kevin and Dwight and I didn’t think that was a good price. We thought the board might be wrong. They didn’t know that the ad volumes had already turned up.

Paparo: David Rosenblatt became CEO. He quickly made some hard decisions. He sold off the mail division, closed Abacus in Europe and sold Abacus, sold marketing automation. The product team was a lean mean machine. There was accountability and process.

Bender: Hellman & Friedman came in and clarified the lines for each of the component parts. They focused on the ad tech use case. They were clear.

Gone was the 33rd Street pleasure palace, aka “Click City,” replaced by a less lavish space on Eighth Avenue. Abacus was sold to Epsilon. DoubleClick was shopped and Google jumped. Its expansion from search into display was not beloved by all. Microsoft objected before the Senate Subcommittee on Antitrust that “this acquisition would give a single firm exclusive control over the largest database of information on individual online behavior the world has ever known.”

But the deal closed in April 2007, with Google paying $3.1 billion in cash. By any measure, it was a very smart move, expanding Google’s power in programmatic advertising. It also domesticated the demon.

Paparo: DoubleClick was on the eighth floor of the Eighth Avenue location, and Google was coincidentally in the same building on the fourth floor. When the acquisition happened, the first thing they did was wheel in truckloads of snacks. That was their first impact.

Bender: Google operates like a small company, even though it isn’t a small company.

Paparo: Everything changed under Google. It had a big impact on the company’s culture. It was more engineering-driven but more whimsical. Less business-oriented. Google cut a quarter to a third of the jobs, and they were unwilling to allow legacy tech to survive.

Bender: I brought a couple ideas to strategic planning to fund, including yield management and the ad exchange. The exchange was ultimately brought to fruition by Scott Spencer and Michael Rubenstein and became AdX.

Paparo: There were no crazy parties. No shenanigans. People had kids. The expense accounts were not like they were before the crash. But our class made a lot more money. The early guys had more fun, but they ended up with stock that was worthless.

O’Connor: It is viewed as one of the most successful M&A deals ever. I’m happy it lasted so long and continues today under the Google brand.

Bender: It’s bittersweet for me to say goodbye to the DoubleClick name after 20 years. But it delivered on the mission of becoming the operating system for advertising.

Hello, World!

What is Paleo Ad Tech? It’s quite simple, really: a passion project by a couple of people who work at a big software company and come from the world of ad agencies, creatives and dreamers. The topic: the history of advertising technology from 1995 to 2015 — aka the glory years. NetScape, DoubleClick, AdECN, Invite Media, Yahoo! and ad networks and the birth of RTB and so on. Big players like Microsoft and Google facing off over ad servers (remember Atlas?) … and smaller players like Undertone and Interclick and Visible Measures … a heroic time of unfettered innovation similar to the early days of railroads, when anyone could lay down track and buy an engine and there wasn’t even a standard width. Amazing times that, each year, as the coverlet of time draws over us, become less vivid and abundant.

We’d like to preserve those years, as historians of technology and observers of the mighty entrepreneur. We also just like advertising and think it’s unjustly deprecated but those who do not understand its role in commerce — and who’ve never worked in that beautiful maelstrom of art, science and sales that is the modern ad business.

Each week, we will interview a person who participated in some way in the ad tech industry during those years: entrepreneurs, data jockeys, salespeople, suits, engineers. Anyone who will talk to us. We’d love to hear your stories.

If you’d like to share, please email Marty: martykihn at gmail dot com.