In 2001, a visitor to a website using ZEDO could do something rather unusual: choose the advertising. A little menu offered computing, cars, travel, fashion and other categories. Pick one, and the system remembered. On participating sites, the reader’s preference helped determine which ads appeared. In an industry busy learning how to guess what people wanted, here was a company prepared to ask.
- ZEDO sold publishers the machinery to deliver, target and manage digital ads.
- ZINC added a way for agencies and brands to buy premium advertising inventory.
- Discovery acquired ZEDO’s assets and technology in December 2021 for its streaming advertising stack.
The wrinkle was supply. If a publisher had not sold enough ads in the chosen category, the reader received general advertising instead. Preference could organize a market; it could not conjure buyers. Launched by Roy de Souza in 1999, ZEDO was introducing this idea as the dot-com collapse thinned the advertising business. Timing made an elegant proposition considerably harder to sell.
01 / The bill for an empty shelf
To understand ZEDO, start with the publisher’s unglamorous problem. Someone must schedule campaigns, select advertisements, deliver them to browsers and report what happened. An ad server does that work. The publisher may own the audience, but turning that audience into reliable revenue requires machinery, operating decisions and buyers.
In January 2002, ZEDO offered to serve unsold inventory free. Publishers commonly filled those spaces with house ads while still carrying ad-serving costs. ZEDO argued that its architecture made a different arrangement possible: charge when the publisher was earning from the advertising. It was a practical response to a bruised market, with the vendor’s bill more closely tied to its customer’s fortunes.
The underlying business was software and services for other businesses. In a 2007 interview, de Souza put typical serving rates around one to five cents per thousand impressions, depending on scale. Advanced sequencing and behavioral targeting carried higher rates and monthly minimums. Those were historical indications, but they reveal the economics: tiny unit prices become consequential when a site serves enormous volumes.
De Souza’s indicative range for ad serving. Media purchases and advanced features were separate considerations.
02 / Asking gave way to observing
By 2007, de Souza was describing a different targeting philosophy. ZEDO had begun with self-selected interests; now he favored learning from what people actually did. That mattered particularly on sprawling sites where the page itself offered little clue about a visitor’s shopping intentions. Behavior supplied a signal that a generic page could not.
This was a change in judgment, not proof that the original menu had simply failed. The founder explicitly described observed behavior as the better approach. The trade-off was clear: less effort for the reader, more inference inside the advertising system. ZEDO’s story contains both ideas, which makes a tidy origin myth rather difficult.
03 / The customer was teaching the company
In 2010, ZEDO’s product list read like a publisher’s collection of daily irritations: network optimization, self-service buying for smaller advertisers, inventory forecasting, mobile delivery and distinctive rich-media formats. De Souza named Alexa.com, Friendster and CBS among its customers. The expertise lay in understanding the operational details between having advertising space and getting paid for it.
He also offered an unusually useful admission. He should have concentrated on premium publishers earlier, he said, and hired senior publishing executives to understand their problems better. Technical capability had not automatically supplied customer knowledge. For another software founder, that is the detail to underline: bring the customer’s operating experience into the room before expanding the feature list.
“I should have focused the company solely on premium publishers much earlier.”
Roy de Souza / 2010
04 / Then the plumbing became a shop
Publishers eventually asked whether ZEDO could help sell the ads, too. The result was ZINC. In 2011, de Souza described an exchange aimed at premium news inventory, with buyers able to inspect availability, prices and audience information before purchasing. ZEDO supplied technology to websites; ZINC sold advertising space to advertisers. The distinction explains why one company could occupy several boxes on an advertising-market diagram.
Its Indian relationships included publishers such as Hindustan Times and NDTV. De Souza reported roughly 50 live Indian sites in October 2011 and said ZINC charged a commission on purchases. The alternative included networks arranging inventory through sales teams and other exchanges, including DoubleClick and OpenX. ZINC’s pitch emphasized premium placements, buying transparency and a more direct route to supply.
05 / A chance to be seen
Premium placement introduces another awkward question: did the ad actually enter the reader’s view? ZEDO’s InView Slider addressed that concern through format design. A 2012 Comscore announcement included ZEDO president Paul Prior’s statement that the format consistently registered 99% viewable impressions under Comscore measurement. It was a claim about a particular format and measurement context, rather than every advertisement ZEDO served.
Visibility is a useful threshold. It does not establish that someone watched attentively, liked the creative or bought the product. The distinction limits the lesson a marketer can copy. Better placement can remove wasted opportunities; it cannot rescue an irrelevant offer. Likewise, premium inventory only supports a premium business when advertisers value the audience enough to pay for it.
06 / Discovery wanted the machinery
In December 2021, Discovery acquired ZEDO’s assets, technology and intellectual property. The announcement identified supply-side platform and real-time bidding capabilities, with employees in India and the United States joining Discovery. The technology was intended to work alongside Discovery’s video platform and server-side ad insertion capabilities, which it had strengthened through its earlier AdSparx asset purchase.
For Discovery, the attraction was owning more of the systems that sell advertising across its direct-to-consumer services. For readers of ZEDO’s history, the transaction supplies a revealing ending. A company that once asked visitors to pick an advertising category had accumulated infrastructure valuable to a streaming business. The recurring opportunity was remarkably concrete: understand where a publisher loses money, then build something that removes the obstruction.