Business Model: Long-Tail Keyword Matching

Digital Media was one of the first large-scale attempts at monetizing the long-tail of Google searches. In their S-1, they claim their "proprietary algorithms and processes" combined with a studio of freelance content creators allowed them to identify and capture opportunities in search.

This was no small operation either. As of June 30, 2010, Digital Media websites comprised the 17th largest web property in the US. 86 million unique visitors and 550 million page views globally. They had 10,000 content creators pumping out a daily average of 5,700 articles and videos. They published over 500,000 pieces of content each quarter!

When someone entered a keyword on Google, a Demand Media property displayed near the top. These keywords were under topics that were easy to research. Lots of "How to..." and "Best...". They would take existing information online and repackage into search-friendly articles and videos.

This isn't all bad - these resources are needed online. Someone has to build easy-to-read articles on useful information. And the people who create this content should be compensated. The issue arose when other content creators - who were willing to spend more time and energy providing a more useful article or video - ranked higher for these keywords.

Problems With Content Farming

Demand Media had the right idea to fill customer demand where it originated. For many keywords, there wasn't strong content on the web. However, part of their strategy was "gaming" Google's algorithms to rank higher. By listing the keywords (and words similar to them) in the domain, title and body of their content, Demand Media properties ranked well.

Google's goal, however, was to provide the best information for the intent of each query. As you can imagine, Google didn't like it when sub-optimal answers were being provided to questions. By measuring things like when a user came back to Google or if they tried another search, Google could determine which websites were answering questions.

For a while Demand Media had the best content for queries. But as with any arbitrage, competition enters to create a more perfect market. For valuable keywords, websites created better content and outranked Demand Media properties. They even knew this might happen:

The game changed from who created content at all to who created the best content. Despite their claims, Demand Media was a high-quantity content creator, not necessarily a high-quality content creator. In June 2010, when Demand Media went public, there were approximately 207 million websites online. Now there are over 1.7 billion. As the volume of websites increased, the competition for search results increased.

Additionally, some changes with Google's algorithms directly targeted content farms like Digital Media. In February 2011, Google released a new algorithm called Panda. Panda targeted thin sites that contained large amounts of advertising.

The other update that hurt them was the Knowledge Graph, launched in May 2012. The Knowledge Graph scraped web content to provide the best answer to the user directly on Google. If the Knowledge Graph was successful, users didn't need to go to another site because Google answers their question before they need to click anywhere else.

As Panda and the Knowledge Graph were implemented, several of Demand Media's sites took a hit. Lots of their content was short answers to quick questions. Content creators were often paid $15-$20 to write articles. It showed.

eHow Example

To explain why lots of Demand Media content lost traffic, let's take a look at eHow. EHow is one of Demand Media's largest properties. On their homepage today, they give a sample of recent articles:

However, very few people find the articles by going to eHow.com. Instead they find them through search engines. So let's see if we can find these. First, "pineapple casserole recipe". Notice how eHow isn't in the top results.

As you can see, websites like Food Network and Southern Living rank at the top for a recipe query. Through several measures (traffic, backlinks, length, related content quality, etc.) Google has determined these are the best content sources to match the keywords. And this makes sense: I'd bet Paula Deen makes a better pineapple casserole than eHow.

For another example, I Googled "How to make your own fabric whitener". In this case, eHow has the top spot.

The tough part here is the snippet. This snippet was created by the Google Knowledge Graph. In scenarios where a 1-2 sentence answer suffices, the user doesn't click-through to the links. Since they found their answer on Google, there's no need to keep searching. The user wins (question answered), Google wins (happy user using Google) but the content site loses (no click which means no visitor for ads). [Note: In this specific case, the user may dive deeper and click on the link, but you get my point].

This eHow example is a long way of saying that Google has a strong incentive to provide the best match between the intent of the search query and the speed and quality of content to match the query. If content creators are maximizing anything else besides providing the best answer to the intent of the query, they are going to have a bad time in the long-run.

Capitalized Content Costs

Besides the issues with content strategy and business model, Demand Media also used wacky accounting to hide profitability issues.

Demand Media amortized their content production costs over five years. So if a creator was paid $50,000 for a year of content creation, they would capitalize that cost and only incur ~$10,000 for that year.

Other publishers don't amortize their editorial costs over the useful life of the content. Tastes change - who knows what we will be consuming in five years? And if any business did have a reason to amortize editorial costs, it probably wouldn't be the type of content Demand Media created.

Own Your Distribution

Even if Demand Media had built a successful content machine, they still had an ad-driven business. And that's a tough business to be in.

The fundamental problem with the business model is the company is at odds with it's user. The company wants to maximize ad revenue. This means placing ads in more disruptive and clickable locations. However, this hurts the user experience. This tension between the advertiser's interest and the user's interest is at the core of why ad-based businesses are really hard.

The story continued downward for Demand Media, driven by lower page views and RPM:

In August 2014 the company overwent a major overhaul. They spun off the domain name registrar segment as a separate entity called Rightside (eventually bought by Donuts Inc). They also purchased Saatchi Art and named Sean Moriarty CEO. They focused on building marketplaces along-side community-focused media properties.

The moral of the story is to own your distribution. 40% of Demand Media's traffic came from search engines. Their valuation eroded when Google changed it's search experience to prioritize the user. If you own your distribution, especially if it's opt-in like email, you know you can always reach your audience. Plus, you know that they actually want to hear your message.

‍