Roy de Souza started ZEDO in 1999, a year when an internet business could acquire a very promising future before acquiring much of a business. The optimism was real. So was the crash that followed. His company had to learn a less fashionable skill: staying open when the enthusiasm had left the room.
By December 2021, Discovery had acquired ZEDO’s assets, technology and intellectual property, with employees in India and the United States joining the buyer. Between those dates lies the part of entrepreneurship that rarely fits into a celebratory announcement. De Souza had spent years working on the machinery behind online advertising, then explaining why that machinery should produce something people might actually tolerate.
He was selling a proposition to publishers as much as to advertisers. A website could attract an audience and still struggle to turn that attention into revenue. The technical question was which advertisement to show. The commercial question was how to make the answer worth more. Anyone who has ever fled a page beneath an avalanche of advertisements will appreciate that the reader had a question, too.
Looking back in 2011, de Souza described the early optimism plainly: “It was a good time to start the company”. Then the market collapsed. His account of what followed was equally practical: the company learned to keep going. There is a certain economy to that description. A bust can be an expensive business school, especially when the tuition is payable immediately.
An engineer learns the checkout
His route to internet advertising passed through engineering and commercial strategy. He studied engineering and economics at St Anne’s College, Oxford, and earned an MBA at Northwestern’s Kellogg School of Management. At Kellogg, he helped establish the TechVenture class. The combination gives his later career a useful point of reference: the mechanism mattered, but so did the market for it.
Before ZEDO, he worked in product management at Zip2, the internet city-guide business associated with Elon Musk. His responsibilities included bringing transaction capabilities to thousands of merchants, with work spanning website design, testing, pricing and software suppliers. It was an education in the distance between putting a business online and making the online business function.
The connection to Musk is an attractive biographical detail. The work itself is more revealing. Pricing, testing and vendor selection have little of the glamour of an internet founding story. They are, however, the places where an appealing proposition becomes a usable product. De Souza’s subsequent interviews repeatedly return to this territory: how the system works, who pays, and why a customer would choose it.
India became part of the engine
When the internet advertising market contracted in 2001, ZEDO built its development team in India to reduce costs. An emergency decision became a lasting part of the company’s geography. India supplied engineering talent and also became a customer market. Keeping offices connected across that distance introduced a different sort of systems problem, one that could not be solved entirely with software.
A 2011 team exercise offered an unusually personal response. Employees drew the high and low points of their lives on timelines and explained them to colleagues. De Souza went first. He described the difficulty of moving from Goa to England as a child, when his accent made speaking in class uncomfortable. The founder put his own awkward years on the chart before asking anyone else to do the same.
That small choice gives the exercise its meaning. Corporate candour can sound rather grand until somebody senior volunteers an embarrassment. Here, the person running the company supplied one. It allowed a meeting about colleagues to become a meeting between people, each with a history longer than the job title on the organisation chart.
“We learned how to keep a business going.”Roy de Souza, 2011
A shopping list with social ambitions
There was also Zebo, a consumer venture he launched in 2006. Its organising idea was disarmingly simple: people could list what they owned, list what they wanted, and find others with useful knowledge about those things. The shopping list had been promoted into a social identity. Shoes, cars and electronics could introduce you before you had thought of an opening line.
De Souza noticed that users treated brands as statements about themselves. One person listed a shirt; another made sure to specify a Lacoste shirt. Ownership was a form of self-description. The site also escaped the tidy categories its creators might have expected: members listed broken hearts among their possessions. Commerce had invited human beings in, and human beings had brought the usual complications.
The experiment belongs beside ZEDO rather than in a footnote. Both ventures considered the relationship between people, products and information. One approached it from the publishing infrastructure; the other let consumers describe their desires themselves. A list of belongings could be useful shopping advice, a little performance, or an admission of longing. Sometimes a website gets all three for the price of one form.
An exchange you could understand
By 2011, de Souza was explaining ZINC, ZEDO’s advertising exchange, through a familiar shopping analogy. Agencies would inspect available advertising space and buy it online. The ambition was to make the transaction more legible: buyers should see availability and pricing before committing, while publishers should have a route to buyers prepared to pay for premium placements.
His distinction between an ad network and an exchange was organisational as well as technical. Networks often grew around sales relationships. Exchanges required the software capacity to manage a more complicated transaction. ZEDO’s existing ad-serving infrastructure gave it a starting point. De Souza wanted the company to move closer to the purchase, rather than remain solely the machinery that delivered the advertisement afterward.
The pitch had a human simplicity beneath its vocabulary. Advertisers wanted to know where their money was going. Publishers wanted better returns. An interface that made those questions easier to answer had a commercial purpose beyond looking modern. Transparency was something a buyer could use before signing, rather than a consoling adjective supplied after the bill.
In India, he also argued for the potential of regional-language publishers. English-language audiences were only part of the opportunity. He pointed to publications including Dainik Bhaskar and Malayala Manorama, and described a commitment to customer support and customisation. It was a view of the internet in which growth would arrive through the habits and languages people already had.
The customer’s inconvenient habits
One detail from his account of ZINC’s strategy deserves its own place. Some Indian publishers sold advertising by the day. The exchange therefore needed to accommodate that practice. De Souza explained that supporting it was a condition of bringing those publishers onto the platform. An elegant system that excluded the customer’s actual way of doing business would have been elegant and empty.
There is a pleasing lack of romance in this example. Strategy becomes the decision to build a particular capability because people need it. The customer does not have to admire the founder’s preferred model. The founder has to understand the customer’s transaction. For a company selling across markets, such details can carry more weight than a declaration that the world is becoming digital.
His later writing extended the argument to the advertising supply chain. In 2015, he described ZEDO’s connection between buying tools, a private network and its publisher ad server as a way to reduce exposure to unsafe intermediaries. Consolidation, initially attractive for time and cost, had become a security argument. He was making a case for knowing the route an advertisement took to the page.
Fewer ads. A harder bargain.
By 2016, de Souza was openly critical of the contest between ad blockers and publishers. A cycle in which readers blocked advertisements and publishers tried to defeat the blocking offered an unhappy arrangement for everyone involved. Readers risked losing free content; publishers risked losing the revenue that supported it; marketers risked losing the audience they wanted to reach.
His proposed bargain involved fewer advertisements on each page, a greater emphasis on brand advertising and higher prices for the individual placement. He also wanted readers to understand the economics of publishing. The interesting part is the concession contained in that proposal: adding more advertising could make the product worse. An advertising executive was asking the industry to exercise restraint.
It was a long way from treating every available impression as an invitation to fill another corner of the screen. His argument recognised that attention has a social dimension. The reader can leave, resent the brand or install a blocker. A business built around showing messages eventually has to take some interest in the experience of receiving them. The audience, inconveniently, gets a vote.
More value per placement.
A proposal for the industry, rather than a reported financial result.
The long route to Discovery
The 2021 Discovery transaction brought that advertising infrastructure into a streaming business. The acquisition included ZEDO’s supply-side platform and real-time bidding capabilities. These are tools for making advertising inventory available and handling automated buying. Discovery also brought ZEDO’s US and Indian employees into its organisation, extending the geographic connection that had helped the company endure its early crisis.
De Souza’s explanation at the time returned to the buyer’s experience: advertisers would have a single place to purchase advertising across Discovery’s streaming platforms. The setting had changed from the early commercial web, but the question sounded familiar. Could complicated advertising machinery make the purchase simpler, while giving the owner of the audience a better way to sell?
Today, his public profile places him in Miami, and Save Groundwater Foundation lists him as a director in India. Away from the advertising vocabulary, a podcast biography records two interests: skiing and audiobooks. They are welcome details in a career otherwise crowded with platforms, exchanges and acronyms. An entrepreneur is allowed an interest that does not require a slide explaining the revenue model.
The durable thread in de Souza’s advertising years is a willingness to revise the arrangement. Development moved across borders. A publisher’s daily rate became a product requirement. A business that delivered advertisements argued for fewer of them. The company launched amid the optimism of 1999 eventually carried its technology into streaming. Along the way, keeping the business useful proved a more demanding occupation than keeping the original idea intact.