Blog Post

Reality Check: For Many Publishers, Google Is Much Less Important Than You Think

Arnon Mishkin is a consultant focused on media and internet companies.

The easiest way to clear a room of internet-savvy media executives is to suggest that maybe they should close their sites to crawling and prevent Google (NSDQ: GOOG) from carrying excerpts of their content. “Look, our ability to sell bundled content is dead, ”they argue, “and we have no choice but to allow Google to use our content – and get links back.”

But the data doesn’t fully support that argument. First off, the experience of the mobile web (both the iPhone and now the iPad) shows that the business model in electronic information is actually more about delivering a bundle (an app) and strengthening one’s brand than in trying to get links from search. And even on the PC-based web, despite all the talk about legacy media losing its readership, many of these sites have been quite successful at building a core group of regulars. Their brands are stronger and Google less important than most realize.

Recent studies reported by the American Press Institute and ITZBelden showed that, at least in the newspaper industry, the predominant share of site traffic comes from a fairly small group of readers. In fact, for most newspapers, some 80%-95% of the traffic comes from a group that is typically smaller than the number of print subscribers to the paper.

If that analysis is correct – and I believe it is – then the question is not, “Is Google essential for driving traffic to one’s site?” Rather, it is: “How much of an audience can one build and sustain without Google?” Indeed, if the web audience a content business needs to create is akin to the size of a newspaper subscription list, then building that readership base is a smaller hurdle than most people think.

Google delivers less value – in terms of search-based visitors – than most people, including the managers of many websites, believe. Why? Thanks to all the users who, for whatever reason, tend to type website addresses into the search box instead of into the browser’s address bar.

For national media companies, the most valuable users they get from search engines (i.e. the people who spend the most time on those sites), are looking to go to those specific sites, either by spelling (or misspelling) the site name or by searching for a specific feature of that site. For smaller local media companies, the numbers are even higher — at 65% or more.

And the traffic — the traffic that comes from people searching for the brand of the site — would come regardless of how much of the site’s content (beyond the brand name) was available to the search engine. Only searches based on content itself depend on the sharing of content, or the “Google bargain.”

In other words, these companies, both national and local, do get some benefit from ensuring that their content appears on the search engine – but substantially less than they probably think. The data here is from, and as with all “long tail” data, it tends to be noisy and subject to interpretation. One could argue that for national media companies the number is more like a quarter or a third, but either way, it is clearly substantial.

There are sites, of course, particularly web-only sites that do get a large proportion of their traffic from people looking for specific topics. For example, over 90% of the Christian Science Monitor’s search-linked audience starts with users seeking a specific story/topic. For Gawker, it’s roughly 80%. By contrast, for competitor TMZ, the vast majority of its search traffic and heavy users are people looking specifically for the TMZ domain.

Now, you could use this data to argue that many media companies could close their sites to search and still remain fairly strong. I’m not going that far, but I do think there’s more to that often-ridiculed recommendation than many think. What’s more, if companies want the economics of web sites to more closely resemble the promise of the economics of tablets and smart phones, they ultimately need to figure out how to increase the power of their brand – that is, they need to identify the regular users of their sites and then monetize them either through subscription or an improved advertising proposition.

To do that, they need to consider the following:

1. Are the visitors who come solely from a topic-based search really as hard to monetize as it appears? The answer appears to be yes, based on time they end up sending on the sites once they arrive. And if you factor in the advertising value of the more directed users, that “drive-by” traffic may be even less useful to publishers.

2. What are the opportunities to rethink one’s web strategy to better preserve the power of the original media bundle? Anecdotal data suggests that readers feel more informed and spend more time with e-readers than they do with traditional web sites. That is a function of both design and technology, but should be replicable.

3. How can smaller media companies – particularly the television stations, newspapers and standalone local web-information companies — build their brand and traffic, given the challenges of getting search traffic based on topics alone? These brands may actually retain relatively more power in their niches than the larger, national media companies.

4. How will companies fare that are skilled at SEO but appear to have a less powerful brand than they may want? They will need to determine how best to translate their power to attract topic-searches into a sustainable brand.

5. Are there opportunities to move the content-based search from the independent search engines to the site itself or potentially to some set of multi company consortiums? Prior to seeing this data, I would have said that was an impossible fantasy, but seeing the remaining power of content company brands, it represents something to consider.

4 Responses to “Reality Check: For Many Publishers, Google Is Much Less Important Than You Think”

  1. Roger Black

    Interesting how such a young “industry” is so filled with myths. The reliance on traffic from Google is based on another myth, that content sites can support themselves with ads—if they can just get just enough audience.

    [PS., I’m an art director, but I caught no editing problems, and read all the way through this excellent post.]

  2. Hampton

    Lots of valid points here. (Unfortunately the atrocious editing was significant obstacle to my comprehension.)

    One nitpick about the comment above: “And Google choice of which ad to display is simply based on….which ad pays the most to Google.”

    This is wrong. Google “Google quality score” to see why.”

  3. googlenaut


    Google’s most vocal supporters are the littlest sites who have built their brand via links, which are found on Google, Techmeme, Digg, Reddit, twitter, et al.

    But what we found in analyzing hundreds of thousands of user click thoughs to content across a variety of sites is that Google traffic is some of the worst. There’s lots of clicks, but in general, those visitors don’t stay to read more, and they very infrequently come back.

    On the other hand, the major media outlets already have subscribers, and an audience. Natural, repeat readers are far more valuable than the drive-bys that Google delivers.

    Is it true then, that Google doesn’t know searcher’s intent as well as they’d have us believe?

    For one perspective, look no further than Google’s own ad model: Every click on an ad — $20B a year– is a signal from the user that Google’s natural search results failed to deliver on intent as well as the ad. And Google choice of which ad to display is simply based on….which ad pays the most to Google. Not very fancy. Google /could/ feed back that ad response data to make their search engine smarter. But dramatically improving search would kill their golden-egg laying goose.

    Still disagree? Perhaps you think “traffic” is critical to making money. Please cast a critical eye over to youtube. The worlds largest online ad company with the worlds largest video site, with the most users and user minutes still isn’t profitable after 3 years of trying.

    That’s 350M unique visitors, 75Billion videos served a year and still less profitable than a so-called “old-media, dead trees” magazine. Whoops.

    Media will do well to stop listening to “web saavy” people and get on with building their own business.