Why nobody clicks on your home page links

Posted on March 1, 2009 by Howard Owens

During my last trip to Boston, I asked a friend: "When is the last time you picked up the Boston Globe and read it cover to cover, every story?"

He starred at me blankly, not comprehending the question.

It was a stupid question, because pretty much nobody ever reads a newspaper cover to cover, not even a small newspaper.

We all read a newspaper the same way — we scan, looking for interesting headlines, skimming the leads, looking for something interesting.

Once we find something interesting, we will start to read and maybe even follow the story past the jump, but the vast majority of headlines that pass before our eyes are merely a blur as we hunt and peck for a useful nugget or two.

Yet, some people seem to think that just because a link on a home page exists, it gets clicked.

If you run a newspaper web site and are under the false impression that just because you put a story link up, people will follow the link, I invite you to open your login to Ominture and study the Paths report. You’ll be disappointed in what you will find.

What you will find, unless some sensational story hit that defies the rule, is that not a single story link is among the top-10 paths followed.

What you will find is the vast, vast majority of visitors hit the home page and left. They didn’t click a single link. The next most frequent path, at between 4 percent and 8 percent of your visitors, will be home page to obituaries. The third most popular path will be home page, obituaries, home page and then exit.

The rest of your top 10 paths will round out with home page to another section front and then exit — meaning, still not a top 10 path that leads to a story click, not even home page, sports section, story link.

When you do get to a home page-to-story-link path, that path will represent little more than 1 percent of your site traffic.

Before you start blaming your site design for this lack of story traffic, stop again and think about how you read a newspaper.

People go to your home page not to find stories to read, but to harvest headlines on the off chance one or two of them will be of sufficient interest for a click.

That’s one reason newspaper.coms are foolish to let aggregation sites such as Topix display all of their headlines and leads.

Topix is in the business of creating a substitute home page for your community news.

By aggregating all of your content, as well as other media covering your town, they are aiming to create an experience for users that says, "You don’t need to visit all of these other sites. We’re all you need. We’ve got all of the headlines (which you will only scan) and free classifieds, to boot (not that Topix free classifieds seem to get much traction).

At GateHouse Media we asked Topix to stop aggregating our content because we couldn’t figure out what value we derived from Topix trying to steal our audience. It would have been different if Topix actually generated traffic for our sites, but referrers from Topix never rose much above 1 percent of our overall traffic.

Some would argue that Topix is paying for its headlines and leads by the traffic it generates, but if it’s not generating much traffic how do you measure whether it’s hurting more than helping?

Compare Topix, however, to a site like Google News.

Google News drives a significant amount of traffic to news sites. Why? Because it has one primary purpose: to drive traffic to news sites. It’s a click-away site, meaning Google believes the greatest value it provides its users is to serve up links worthy of a click.

My bet is that most of the clicks driven by Google News are derived from search, not from the automated aggregation pages. People click on headlines when they express a specific intention through search to find a particular story.

As I’ve said before, the web is intention driven. If your home page is designed to meet the intention of headline skimmers, that’s going to be the majority of your audience. But if your home page is designed to get people into your stories, like a blog does, then you will design your site accordingly.

Think of how you read a newspaper and don’t be surprised that few people click on your headline links. Think about how you want people to use your web site, what intention-driven mindset you want to satisfy, and design your web site accordingly.

Why home page ads may be more valuable than story page ads

Posted on March 1, 2009 by Howard Owens

All web activity is intention driven.

People visit web pages, whether arriving via search, a link or a bookmark with a specific intention. That intention might be to read a specific story, see what’s on sale, scan headlines or connect with a friend.

How well a web page helps a user satisfy that intention determines whether a user will return to that page or recommend it to others.

The page may not efficiently satisfy a user’s intention — the web world is full of poorly designed pages that survive by providing a marginal benefit to users, newspaper.com sites chief among them — but so long as the user is free to focus on that intention devoid of distractions or unexpected interruptions, the user experience will be OK.

Much has been made of eye track studies that demonstrate banner blindness. What’s interesting is the only "banner blindness" eye track reports I’ve been able to find demonstrate banner blindness on story pages.

I’ve never seen such a study — and if you have, please let me know — on a newspaper.com home page.

The banner blindness studies support, I think, the proposition that user behavior on the web is intention driven. When a user clicks on a link — whether from aggregator, search engine, blog or newspaper.com home page, the user has expressed an intention to read a particular story or post. The user is solely focused on that task, so she ignores the banners.

But what is the intention of a user visiting a newspaper.com home page?

I do not have available to me an eye track study to support my theory, but I do have years of experience studying heat maps of user behavior on home pages in Ventura, Bakersfield and GateHouse Media, and I believe the user intention is to scan the newspaper.com home page looking for something interesting.

Notice, I didn’t say "something interesting to click on." Just "something interesting."

Users visit a newspaper.com home page not so much because they want to dive deeper into the site, but because they want to see what is new.

We can debate whether the typical newspaper.com is doing well at satisfying that intention, or more importantly, whether that is the right intention to meet, but I believe that is the typical user intention.

Most such well-intentioned users are most likely looking for the latest news, or other new content, but I would contend that a scanning user is a user who is more likely to take in the full breadth of the home page — they’ll see your top nav links, your promos for your special features and, most importantly, your home page advertising.

This is why it’s probably a mistake for newspapers not to put more advertising on their home pages. The home page audience is more likely to notice a home page ad than an story page audience. (I know there are studies that contradict this theory, that more ads on the home page lead to less effective ads, but I don’t believe this proposition has been fully and fairly studied at the community news level, where local ads tend to be highly relevant to local users.)

And it’s also why newspaper publishers should think about how to get more visitors to the home page. That’s where the money is, and that’s best vehicle for generating audience growth.

Conversely, story pages need to be parred down to the essentials. Banner ads on story pages are a waste. Contextual ads might have some value, but the best move a publisher can make with story pages is use single-focus pages as a vehicle for promoting other content.

By visiting a story page, a user has expressed at least a marginal interest in the content you have available. Use the story page to present more content, be it top headlines, most e-mailed stories or "related stories."

I’ve seen page views increased by 10 percent with the introduction of a pretty low-tech "related content" widget.

Giving users more content choices on a content page works — more advertising choices, not so much.

Your goal as a newspaper.com publisher is to increase user loyalty. Your ideal user visits multiple times per day, ideally by constantly refreshing your home page to see what is new. The more you can entice the occasional visitor into reading your content, the more likely that user is to become a frequent home page visitor.

If your advertising is highly relevant to that user, he is more like to take notice and also be inclined to support the businesses that support your news operations. It becomes a virtuous circle.

When designing your web strategy, think constantly of user intention. Ask, how are people going to use this page? Then design your page strategy around that intention so that both the user’s consumer needs and your business needs are satisfied.

The myth of the deep link

Posted on March 1, 2009 by Howard Owens

Not all links are created equal.

Not all links generate traffic.

Some links are downright harmful.

Yet, to a large class of digerati, the link is a fetish.

Woe to he who questions the value of a link.

I’m sorry to state the truth, but just because you create a link doesn’t mean a single user will click on it.

Even Google doesn’t treat all links equally. Internal and external links appear to get weighted differently, and the better the Page Rank of a site, the more valuable are its external links. For Google, too, the words associated with a link play a role in how its algorithm evaluates the link.

Not all links are created equal.

And some links can be downright harmful.

The first time my blog was hacked, some spammer filled it with redirects and then linked to it from a hundred other blogs. Google immediately punished my site by lowering its Page Rank. HowardOwens.com went from the first search result for "Howard Owens" to a single interior story link on the fifth page of search results. I was dead to Google for about three days. (Thanks to Matt Cutts for resurrecting me.)

Needless to say, I wasn’t happy with those 100 unwanted links.

A link can also be harmful when it is associated with words that misrepresent the content of a site. This can harm search results or leave people with a false impression of the site being linked to.

A link from a mega site like Fark might be fun, but it isn’t necessarily helpful to a local publisher. Some publishers will complain about the non-monetizable traffic, but bandwidth sucking links isn’t the real harm. The real harm is for publishers who sell ads on a CPM basis.

A local advertiser, on a CPM model, can see an entire month’s worth of bought-and-paid-for inventory served up to a non-local audience in a matter of hours based on a Fark link. And the Publisher is left serving up less valuable remnant ads for the balance of the month, and the advertiser is left wondering why his ad stopped appearing on the site.

Modern ad serving software has methods to help account for such spikes in traffic, but such balancing isn’t perfect and some impressions are wasted. Mainly, the example still demonstrates that not all links are created equal.

And depending on the context, a link might act as a substitute for actually visiting the site receiving the link.

If a headline and lead, for example, tells a user all he or she needs to know about a particular news story, why would that person click on that link? In that scenario, you can drew one of two conclusions: Either the story wasn’t sufficiently compelling that the non-visiting user probably wouldn’t have gone to your site to find such a story anyway, or the user will decide, "well, now I don’t need to visit that site because I already know all I want to know about the news."

One result is neutral, the other result is harmful.

None of this is to say I don’t believe in and support deep linking, or linking of any kind.

Back in 1996 or so when the first arguments over deep linking emerged, somebody on Steve Outing’s old Online-News e-mail discussion list pointed out the value of linking, of networked sharing, by using this metaphor: when the water level increases, all boats rise.

I still believe in all boats rising, but I’m also not interested in making a fetish out of the link.

Any professional charged with growing a web business needs to make calculated observations about the benefit or harm of any web practice and decide for himself whether a particular practice or belief is going to benefit the long-term viability of the business.

When you blindly follow the herd, you’re not doing your job. I’m sure a guy like Eric Schmidt at Google would agree. That’s "What Google Would Do" — encourage you to make your own evaluations and observations.

Linking is one of those issues that should be carefully considered so that you ensure your linking practices and policies are a benefit to your business organization and not a detriment.

Not all links are created equal.

The myth of multiple gateways into a news site

Posted on March 1, 2009 by Howard Owens

If you run a small town, local community newspaper site, the most important page on your server is your home page.

Take a look at your stats: More than 50 percent, and maybe as much 70 percent of your Web traffic flows to your home page.

Now, some people who think they understand SEO might step forward and say, "Well, you’re just not correctly optimizing your content for Google."

I say, those people don’t know what the hell they’re talking about.

It takes a little thought, but if you look at the typical small town newspaper web site, you’ll understand that the content of such sites serves a narrowly focused audience — people who live in that town (and a few stragglers who once lived there).

Now, the occasional story might arise that generates global interest, but on a day-in, day-out basis the content a local newsroom produces is of merely parochial interest.

No matter how well your site is optimized, if few people are searching for Joe Bubba’s DUI arrest, that story isn’t going to show up in Google. It’s the "tree falling in the forest problem." Even if your story is indexed and highly optimized, if nobody ever enters search terms that brings the story to the surface, the story might as well not exist in Google.

And being a small town site, there are likely few if any bloggers who are likely to link to the Joe Bubba story. I’m sorry, but unless Joe Bubba is somehow tied to Newt Gingrich, neither Instapundit nor Daily Kos is going to link to his arrest, no matter how shocking it is back home.

SEO has its place, but it doesn’t negate the importance of a newspaper.com’s home page.

I don’t have the documents in front of me (and it’s not online as far as I know), but Greg Harmon of Belden once showed me research that indicated about 70 percent of the traffic of a small-circ newspaper.com came from visitors within that paper’s DMA. In my own observations of traffic patterns in Ventura, Bakersfield, with GateHouse Media and running The Batavian, I would say that’s roughly true.

And it makes sense. Again, the vast majority of content produced by a local newspaper is of purely parochial interest. If your in Los Angeles, you are not going to have much cause to visit the Web site for the Freeport Journal-Standard, unless you were from Freeport, Ill. or had family there.

Local news sites live or die on how well they meet the needs of a local audience.

The same cannot be said for major metro sites, and perhaps this is where some of the confusion comes from on this topic. The bigger the newspaper, the more bloggers there are who follow it’s content, the more often it covers stories of a googable interest, and the bigger its global audience.

This is certainly true of sites such as the New York Times, CNN, Boston Globe and Chicago Tribune, for example. I’ve heard, but have not seen the actual stats, that as much as 70 percent of a major metro’s traffic flows to interior pages.

Expecting that much interior traffic for a small town site is like hoping a banker will turn down his bonus. It’s just not going to happen.

This is why small-circ newspaper publishers need to protect their home pages like Obama clings to his Blackberry. It is the key to revenue and audience growth.

Most local publishers have piss-poor home pages, but that’s an issue for another blog post. But even the worst newspaper.com home page is more valuable than the aggregate of all the internal story pages. In part, that’s true, because the home page is the only page most of that 70 percent local audience will typically visit. But, again, that’s a topic for another blog post.