The Search Engine That Followed the Money

Dave Sifry wanted to know when somebody on the internet was talking about him. In 2002, that was an awkward question to answer. Blogs published quickly, linked promiscuously and vanished into yesterday almost as fast. So he built Technorati, a search engine for that unruly conversation. For a while, the company was the blogosphere's clock, map and social register. If you wrote a blog, Technorati could tell you who had noticed.

The tale is sometimes told as a clean pivot. It was messier and more interesting. Technorati changed its answer to three separate questions: who was its customer, what did that customer need, and where could the company collect a fee? The answers moved from the curious reader, to the working blogger, to the publisher's ad operations desk. The skill underneath - watching a noisy market in real time - stayed surprisingly similar.

The web needed a gossip columnist

Technorati's first job was discovery. It indexed blog posts, registered incoming links and made them searchable while a story was still warm. Its authority measure turned links into a kind of public status score. Bloggers used it to track who cited them; readers used it to find voices beyond the big portals; marketers could listen to what people were saying about a film or product. In the mid-2000s, those were practical questions, not nostalgia.

The company did have a commercial instinct early on. It sold sponsored links and licensing deals, including arrangements with the Associated Press and Paramount Pictures reported in 2007. But search was the public face. Technorati's 2008 State of the Blogosphere said its index had tracked more than 133 million blogs since 2002. The qualification matters: that was a cumulative count, not a tally of active writers. Even so, it conveyed the size of the organizing problem.

Different dates, different measures. Each number describes a separate stage of the business.

There was a catch. The better Technorati made blog search look, the more inviting that category became to a company with a vastly larger distribution pipe. Google launched Blog Search in 2005. By late 2006, traffic data cited by Inc. showed Google ahead in US blog-search visits. Technorati's first clear failure was not a lack of invention. It was the loss of a habit: a user could get an adequate answer from a search box already near at hand.

The company also had to keep a fast-growing index fast. Contemporary reporting described sluggishness, downtime and missed posts. Readers might forgive an imperfect archive. They are less patient with a service that promises to tell them what happened five minutes ago. Technorati kept adding features and experimented with broader audiences, but its specialist edge was becoming harder to defend.

The customer was sitting on the other side of the screen

The pivot began when Technorati treated its index and publisher relationships as commercial infrastructure. It developed Technorati Media, an ad network that put advertiser demand on participating sites. Instead of simply showing a blogger how many people linked to her, it could help her earn money from the audience she had assembled. Advertisers bought reach; publishers supplied inventory; Technorati sat between them.

That network business had its own limit. Selling ads through a network meant asking publishers to trust one intermediary's buyers and sales operation. Programmatic advertising offered another route: let more buyers compete for each impression. The old method often formed a waterfall, with one demand partner getting a turn before the next. A buyer farther down the line could be willing to pay more, yet never get a proper chance. Sequential convenience could leave money on the table.

“We see this continuing down the path of moving away from the network model and into SaaS.”Shani Higgins, Technorati CEO, 2015

Technorati's answer was Contango, a platform for publishers to connect demand sources, and SmartWrapper, the 2015 product that managed header bidding. Header bidding invited several partners to bid before the publisher's ad server chose a winner. SmartWrapper put partner tags under one tag and gave publishers a shared place to manage integrations, reporting and slow bidders. The difference was operational. A publisher did not need to hand its whole ad business to a single network to use the tool.

What changed: buyers could be asked at roughly the same point in the process, instead of waiting in a fixed queue.

There was a price for this more democratic auction, though public materials do not disclose Technorati's standard fees. SmartWrapper was licensed as software. Publishers still paid in engineering time, vendor coordination and possible page latency. The product itself tried to reduce those costs through one integration, throttling of slow partners and centralized analytics. Its sales pitch made most sense for publishers large enough to have multiple demand partners, but lean enough to resent maintaining each connection by hand. A small site with one serious buyer would have less reason to add another auction layer.

A company can change products without changing its eye

Technorati's last independent product looks far from blog search until one notices the common problem. Search compared a flood of posts and links; SmartWrapper compared a flood of bids. Both asked what matters now, and both needed speed. The company knew publishers, their traffic and their complaints. That knowledge helped it move from a consumer-facing search habit to a business-to-business tool.

There is a practical idea here for anyone building in a crowded market. Watch what your users do after your product delivers its answer. A blogger did not merely want a score. She wanted readers and, often, income. A publisher did not merely want a dashboard. It wanted a higher price for its next impression without slowing the page. Technorati followed that chain. The move works only when the next problem is painful enough to fund a solution and when the company has credible access to the customer who owns it.

It is also easy to overstate the outcome. Technorati had raised about $32.1 million across its funding history according to funding databases. Synacor said the company generated approximately $7 million in 2015 revenue, served more than 1,000 publishers and added more than 100 million monthly unique visitors to its advertising reach. Yet Synacor's SEC filing records a $3 million cash price for substantially all Technorati assets in February 2016. An asset sale price is not a neat measure of investor returns, but the gap between capital raised and sale price is striking.

For Synacor, the attraction was specific: a publisher network, advertising technology and a tool that fitted its monetization business. For Technorati, the sale ended its independent operations. Its original service had already stopped indexing blogs in 2014. The name that once meant “who is talking?” finished attached to the more prosaic question “who is bidding?” That may be the most accurate portrait of the open web's adolescence: attention became inventory, and the people who learned to count it learned to sell it.